Introduction

Success in clinical trials for new drugs is a critical stage for both biotech and pharma companies. The discovery and the development of a new drug is a long and costly process, with high-risk of failure. This process can take over 10 to 15 years, with $1-2 billion average cost.¹

The current failure rate of clinical studies for new drugs, from phase I to final clinical authorization is more than 90%². That’s why pharma companies work on the improvement of probability of success (POS) in clinical trials. This POS is critical for clinical researchers and biopharma investors to evaluate when making scientific and economic decisions.

The main reason of clinical trial failure is the impossibility to identify a real clinical efficacy with a new drug (40-50%)². One way to increase POS is improving the criteria for selecting patients included in the clinical studies. The democratization of companion diagnostic (Cdx) use in clinical studies in oncology is moving in this direction³

The subject of this application note is to discuss how using complementary diagnostics (CoDx), and specifically functional assays, during clinical trials in oncology can improve the success rate. The Oncogramme®, a standardized functional assay, allowing the quantitative evaluation of the anti-cancer capacities of a molecule directly on primary cancer cells (2D culture) extracted from a patient's tumour, is the example that we chose to illustrate our discussion^4–8.

1. Clinical studies in Oncology: A long process with high risk of failure

Drug discovery and development is a more than a decade long and high-risk process. The estimated costs for a clinical approved drug is between 1 and 2 billion dollars. Thus, the clinical trial success is crucial for pharma companies, after years of research, development and preclinical studies. Nevertheless, only one out of ten drug candidates, entered in a clinical study, will finally be clinically approved. It’s important to specify that these 90% of clinical study failure include only drug candidates entered in a clinical study and not all the drug candidates eliminated during preclinical studies².

As explained in an article from Health Network University of Toronto, Clinical trials are divided into phases, to provide key decision points, to continue or stop during the drug development path. Phase III is a crucial step in drug development process. It represents the most costly step and needs a long time before obtaining the results⁹.

An article, from MIT, studied the clinical trial success in terms of disease type, clinical phase, industry or academic sponsor, biomarker presence, lead indication status, and time. For oncology, this study identified 3.4% success rate from phase I to clinical approval.   The same study identified 235 published phase III randomised clinical trials in Oncology from January 1, 2000 to October 31, 2015. The article reported that 62% of oncology phase III clinical trials did not achieve results with statistical significance³.

Another article, from University of Michigan, tried to identify the reason of these 90% rate of clinical trial failure². Four reasons were identified by authors: lack of clinical efficacy (40%–50%), unmanageable toxicity (30%), poor drug-like properties (10%–15%), and lack of commercial needs and poor strategic planning (10%)².

As the main reason for clinical trial failure is lack of clinical efficacy (40%-50%)², a way to decrease the clinical trial failure rate, due to lack of clinical efficacy, is improving the patient selection, in particular in Oncology, where major differences are observed between several cancers from the same location.

Using biomarkers is a possible way to improve the clinical study success rate. On 406,038 entries of clinical trial data for over 21,143 compounds from January 1, 2000 to October 31, 2015 a MIT team identified only 7.1% of all drug development paths that use biomarkers use them in all stages of development. The author furthermore identified that 92.3% of the trials using biomarkers are observed only on or after January 1, 2005³.

The same team highlighted that a biomarker is used in 10.8% of clinical studies in Oncology. The overall rate of an oncology clinical study success is 1.6% of cases. With a biomarker, the success rate is 10.7%. A focus, on phase III and more, identified that without biomarker the success rate is 33.6% and with biomarker 63.6%. So the conclusions of this study is that oncology clinical trials with a patient selection using biomarkers have higher success rates³.

If democratization of CDx, for biomarker detection, in oncology clinical studies, increases the clinical study success rate from less than 2% to more than 10%, the risk of clinical studies failure remains very high. The emergence of a new subfamily of CDx, the Complementary Diagnostics (CoDx), offers new opportunities to improve the selection of patients for their inclusion in clinical studies in oncology and so, increase the success rate.

2. Complementary diagnostic: Can functional assays help in clinical studies in oncology?

Current strategies for cancer treatment clinical studies were developed by demonstrating improved efficacy in patient cohorts defined by similarity of diagnosis. As cancers are complexes and heterogeneous diseases, they require a tailored therapeutic management^7,10. The «personalized medicine» term appeared after observation of unsatisfying patient response rates and survival with new anticancer drugs, including immuno-oncology drugs⁷.

The beginning of personalized medicine in oncology, with a biomarker strategy linked to a Cdx development, was in the 1980s¹¹. This biomarker approach quickly spread for clinical trials and clinical treatments in oncology.

The Cdx definition is not totally the same in terms of regulation. For FDA (USA), it is: A companion diagnostic device can be in vitro diagnostic (IVD) device or an imaging tool that provides information that is essential for the safe and effective use of a corresponding therapeutic product¹².

For IVDR (Europe), “Companion diagnostic” means a device, which is essential for the safe and effective use of a corresponding medicinal product to:
(a) identify, before, and/or during treatment, patients who are most likely to benefit from the corresponding medicinal product; or
(b) identify, before, and/or during treatment, patients likely to be at increased risk of serious adverse reactions as a result of treatment¹³

An article from Evnia ApS tried to describe the principal objectives of a Cdx as follow¹⁴:

(a) Identify the appropriate patient group who are most likely to benefit from a therapeutic product;

(b) Identify the patient groups for which the therapeutic product has been adequately proven safe and effective, allowing for adjustment of treatment to achieve optimal safety;

(c) Predict serious adverse reactions that some patients may present as an outcome of the therapeutic drug used;

(d) Monitor the response to treatment to improve/adjust the dosage scheme and to ensure continued patient safety.

So, by identifying the presence or the absence of a drug target on the patient cancer, Cdx can guide the oncologist toward a specific therapy. In the other hand, Cdx are used by onco-drug development companies during clinical trials to identify populations that can be treated by a drug.

However, two limits quickly appeared to the use of target-based biomarkers, in clinical studies or helping the oncologist to choose the best treatment for the patient:

► The positivity of a biomarker, or the presence of the drug target, does not mean efficiency of the treatment. An inherent problem with target-based biomarkers is that they test only for the presence of the drug target and do not account for resistance mechanisms unrelated to the target¹⁰.

► Another problem is that some targeted drugs, like receptor-kinase inhibitors for example, don’t have only a targeted effect, but generate a wide spectrum activities in the cell, which are not tested for¹⁰.

Another type of diagnostic tool appeared more recently: Complementary diagnostics (CoDX). If Cdx is a rigid process from patient to specific treatment (1 positive biomarker = 1 targeted therapy) CoDX can provide information about the potentially enhanced benefits of receiving a drug. These diagnostic tests do not make a specific drug mandatory, though, and a negative result does not disqualify the linked drug. So, the main difference between Cdx and Codx is the freedom of decision for the physicians regarding the choice of treatment for their patients¹⁵. To summarize, a CoDx is a test that aids in the benefit-risk decision-making about the use of the therapeutic product, where the difference in benefit-risk is clinically meaningful¹⁶. So far, CoDx, like Cdx have been mostly developed in cancer indication and possess a crucial role in personalized medicine¹⁵.

A CoDx, close to patient subtype, namely functional assays is particularly interesting. Functional assays test on a patient’s own cells, upstream of treatment initiation, the arsenal of therapies  available for a specific indication. So, instead of identifying the roots of disease phenotype (biomarker), they capture the final, and, as such clinically relevant, response to a drug produced by the interplay between all biological variables. These tests are the transposition of preclinical in vitro assays, led on selected models to study the response to a drug candidate, to clinically-applicable ex vivo assays. So, functional assays may hence provide a personalized medicine approach for systemic treatments, including chemotherapies, which have not been clinically associated with single or groups of biomarkers yet¹⁵.

In oncology point of view, predictive functional assays and more specifically chemosensitivity and resistance assays (CSRA) are a very promising path to help oncologist on clinical therapy decision and improve individualized and personalized therapy. A crucial point to understand the importance of functional assays for personalized therapy is predictivity of these CoDx. Two meta-analyses studied the predictivity of CSRA. The first one, form Uppsala University (Sweden), published in 2017, analysed 1,835 CSRA individual patient results, from 34 reports, and identified a CSRA 88% specificity (CI: 86%-90%) and 72% specificity (CI: 68%-75%).¹⁰

The second meta-analysis, from Oncomedics (France) and Centre hospitalier universitaire Dupuytren (France), analysed 42 studies and identified, in terms of cancer localisation, a sensitivity between 44.4% and 100%, with a median of 98%. For specificity, in terms of cancer localisation, between 18.2% and 100%, median 77%. To summarize, the correct prediction percentage, identified via this study is between 44.4% and 94.4% with a median value of 77.8%¹⁵.

This high predictivity of CSRA functional assays can give confidence to Oncologists when choosing the therapies to undertake for patients.

For the rest of this application note, we are going to highlight, using Oncogramme® example, how functional assays’ high predictivity can secure a phase III oncology clinical trial by helping inclusion patient selection.

3. The Oncogramme, example of Chemosensitivity and Resistance Functional Assay for metastatic colorectal cancer

Colorectal cancer (CRC) is a major public health concern and the second leading cause of worldwide cancer-related death (880,000 deaths per year)¹⁷. Metastatic colorectal cancer shows a poor survival rate of patients, around 11.4%¹⁸.

Current standards of care are combinations of 5-fluorouracile (5-FU) and folinic acid (FA, also known as leucovorin) with either oxaliplatin (FOLFOX), irinotecan (FOLFIRI), or both (FOLFIRINOX and FOLFOXIRI). These combinations may be supplemented with anti-VEGF (bevacizumab) or anti-EGFR (cetuximab or panitumumab) antibodies, depending on the tumour’s KRAS/NRAS/BRAF mutational status¹⁹. These therapies are associated with known toxicities, whose severity depends on patients’ age or co-morbidities and may worsen their overall condition, hence influencing therapeutic decisions⁷. Relapsing, microsatellite instable (MSI) mCRC patients may also receive immune checkpoint inhibitors nivolumab or pembrolizumab^20,21.

It is important to mention that a patient with metastatic colorectal cancer (CRCm) is expected to have 50% probability to respond to first-lime combination chemotherapy²². That’s why the French company Oncomedics worked to develop a chemosensitivity functional assay, able to determine response profiles to chemotherapeutics agents on CRC, breast and ovarian. Oncogramme® for CRCm is CE-IVD validated.

Briefly, Oncogramme® consists in the measurement of therapy-induced mortality on patient sample primary 2D cultured cancer cells, using fluorescence microscopy. This test is fully standardized, allowing both reliability and a high success rate. When applied to mCRC, the Oncogramme® directly evaluates drug combinations commonly used by oncologists, such as 5-FU, FOLFOX, FOLFIRI, or FOLFIRINOX. A pilot study performed on a cohort of patients with mCRC showed a sensitivity of 84.6% of the assay in predicting tumour response⁷, which is sensibly higher to published literature on CRC chemosensitivity assays²³. A randomized clinical study, including 256 patient with CRCm and 13 French clinical centers is currently in progress⁸.

The Oncogramme® experimental procedure (Fig. 1) takes place in less than 15 days.  This procedure begins on day 1 by the biopsy or the excision of the tumor by a surgeon. The excised tumor is sent to the pathology laboratory, which separates the tumor into several parts, one of which is dedicated to the Oncogramme®. The tumor part is placed in a conservation medium with a proprietary formulation, which preserves the viability of the cells for 48 hours at 4°C, i.e. the time necessary for the delivery of the sample from the hospital to the laboratory. Once in the laboratory, the tumor is dissociated using a proprietary medium, allowing high cell viability (>80%) and preservation of contamination in more than 95% of cases. Subsequently, the tumor cells obtained are cultured in another proprietary medium, allowing the purification of the cell population and its enrichment. After culture, patient cancer cells are brought into contact with the different chemotherapies, and the induced cell death is analysed by a double fluorescence staining system. This is followed by the drafting of a report presenting the standardized response profile of the patient's tumor, thus allowing the practitioner to choose the treatment that will identify as the most effective on the patient cancer cells, with a predictivity of 84%.

 

Figure 1: Overview of the Oncogramme® experimental procedure, from surgery to readout. Viable samples were recovered and processed to obtain primary cultures that were subsequently utilized for realization of the Oncogramme by exposure to chemotherapeutic drugs and cell death analysis. Whole time course was less than 2 weeks

Results of Oncogramme is a diagram showing quantitive sensitivity or resistance rate of the patient tumor cells to the different chemotherapies. On the Fig. 2, we can observe as an example, the results of tumor cells resistant to 5FU + Folic acid and sensitive to Folfirinox, Folfox and Folfiri. Maximum tumor sensitivity is observed with Folirinox.

The crucial point for IVD CRCm Oncogramme is that using this functional test makes it possible to go from less than 50% chance of response to the first line chemotherapy²² to 84% of chance⁷. So, without Oncogramme, the patient has one chance out of two to receive an effective chemotherapy in first line²². Using a therapy identified as effective by Oncogramme, the patient has 84% of chance to have a real effect on his tumor⁷.

 

Figure 2: Example of Oncogramme graphic results

The Oncogramme is based on a standardized technological platform, usable not only for diagnostic but also for research, preclinical and clinical studies, based on 3 pillars:

First, a standardized tumor treatment procedure.

Second, a kit of culture media and proprietary reagents for sending the sample, dissociation and specific culture of primary cancer cells.

And finally, a proprietary algorithm for automatic counting correlated with clinical data^4–7,9.

These three Oncogramme pillars make it possible to obtain more than 80% of predictivity⁷. This technological platform has been developed in Oncomedics for CRC, breast, ovarian, glioblastoma, lung, and can be extended to other cancers.

The next part of this application note is developing how the high predictivity of Oncogramme technological platform can help pharma and biotech companies to secure phase III clinical trials in Oncology.

4. How can Oncogramme secure phase III clinical studies in Oncology? Example for a new drug against metastatic colorectal cancer

As explained previously, Phase III, especially in Oncology, is a crucial step in drug development process, because it represents the most costly step and needs a long time before observing the results. Additionally, phase III compares the new treatment with the standard of care. Regarding this specific point, literature highlighted how much it’s difficult to identify, for a new drug, a statistic difference versus standard of care³. As an example, a study on 1245 clinical trials in oncology between January 1, 2000 to October 31, 2015 highlight a POS for phase III to approval of 35.5% (SE 1.4%)³. Using a biomarker can improve the rate of phase III positivity from 33.6% (SE 1.4%) without biomarker to 63.6% (SE 5.5%) with a biomarker³. However, even with a biomarker, almost 40% of phase III clinical studies fails, a half due to a lack of efficacy².

We are convinced that, due to high predictivity of CSRA functional assays, it is possible to use them to improve the success rate of clinical trial in oncology. For this, we choose an example: Using CRCm Oncogramme to improve patient selection in a phase III clinical trial for a new drug against   metastatic colorectal cancer (Fig. 3).

In a classical clinical study for a new anti-CRCm drug (Fig. 3A), randomized patients are separated in two branches. First branch with a half of the randomized patients, treated with the new drug. The second one, with other half of the patients, treated by standard of care, Folfox or Foliri. If we decide to add 300 randomized patients in the clinical study, 150 patients will receive standard of care (Folfox or FOLFIRI). As previously published, we will observe a first line patient answer in less than 50% of cases²². So, on 150 FOLFOX or FOLFIRI treated patients, NR (number of responding to drug patients) might be 72. For new drug, we will test 150 patients too. To obtain a clinical trial success, the new drug must have more than 50% of positive answer. So, there is a high risks to not obtain statistical difference between new drug results and standard of care results, so a clinical study fail.

Our proposition, to improve the success rate of this clinical trial, is adding an Oncogramme on the new drug branch, before treating patients with the new drug (Fig. 3B). We choose to create a study with 300 randomized patients, 100 patients on standard of care branch, 200 patients on new drug branch. In this new clinical study, the standard of care branch, with FOLFOX and FOLFIRI, stays at less than 50% of responding patients²². If we treat 100 patients, about 48 might have a real drug effect. On the other hand, for the new drug, we can perform Oncogramme on 200 randomized patients. This first step will drop-out all the non-responding to the new drug patients⁷. If effectiveness of new drug is something like 50%, one half of the 200 randomized patients, identified as new drug non sensitive tumors using the Oncogramme, will be dropped-out. For this example, we choose to drop-out 104 non-responding to the new drug patients. So, only the patients identified with a sensitive tumor using Oncogramme, 96 patients in this example,  will receive the treatment with the new drug. As the predictivity of Oncogramme on metastatic colorectal cancer is 84%⁷, we will observe a real effect of the new drug on 84% of treated patients. So, something like 81 responding patients for 96 treated patients. In this case, we can identify a real statistical difference versus standard of care (less than 50% of responding patients for standard of cares VS 84% of responding patients for New Drug).

This example illustrates that Oncogramme drastically improve the chance of success in clinical study by selecting the new drug responding tumors and secure Phase III clinical trial. Other major advantage, using Oncogramme drastically decrease the number of non-responding to the new drug patients that receive the new drug without any effect against their tumor, but with secondary effects of the new drug. In the classical study, 50% of the treated by the new drug patients receive the new drug without anti-tumor effect. Using Oncogramme, only 16% of the treated by the new drug patients receive the new drug without anti-tumor effect.

A

 

Figure 3A: Classical Phase III clinical study for new anti-CRC drug. Phase III clinical study for a new anti-CRC drug including patient selection without CRC Oncogramme.

 

B

Figure 3B: Classical Phase III clinical study for new anti-CRC drug. Phase III clinical study for a new anti-CRC drug including patient selection with CRC Oncogramme.Oncogramme on the new drug branch, before treating patients with the new drug.

Discussion

Oncology clinical study is a long and costly process with a high rate of failure². Since years, the biomarker strategy, via Cdx, tried to decrease the high rate of clinical failures, linked to a lack of clinical efficacy (40%–50% of the clinical trial failure)². This approach, specifically for phase III of oncology clinical studies, succeeded to increase the success rate from 33,6% without biomarker to 63,6% with biomarker³. However, the phase III failure rate stays high.

In this application note, we are proposing a new approach to improve the oncology clinical study success rate: Using a CoDx CSRA functional assay to improve the patient selection during phase III of clinical studies in oncology. 

► The high predictivity (more than 80%)¹⁵ of CSRA functional assays make it possible to:

1. Securize and increase success rate of phase III clinical study in oncology by selecting, with a more than 80% predictivity the responding to a new drug patients.
2. Decrease the number of patients, included in a phase III clinical trial in oncology, that receive the new drug as treatment without any anti-tumour effect.
3. Increase the valorisation of the new anti-tumor drug with an associated CoDx
4. Work on drug repositioning by identifying new responding population

► We are convinced that Oncogramme technology platform is the best CSRA     candidate for your phase III clinical study in oncology because:

1. Oncogramme is a standardized and easy to develop technology platform linked to clinical
2.  High predictivity of Oncogramme results (more than 84% of predictivity)⁷
3. Oncogramme can predict anti-tumor activity not only for chemotherapies but for all drugs with a direct effect on tumor cells (immunotherapy, ADC,…)
4. Oncogramme can predict synergic effect like Chemotherapy + Chemotherapy, or Chemotherapy + Immunotherapy.
5. Ready for IVD functional assay: Colorectal cancer (IVD); Breast and Ovarian (ready for IVD); Lung, Glioblastoma, Prostate,… (In progress)
6. Long experience of working with biotech and pharma companies
7. Research and development perform internally in Oncomedics laboratories. The Oncogramme experiments perform in a major European clinical diagnostic laboratory.

References: