The drug discovery process is complex and protracted, and requires drug development companies and contract research organizations (CROs) to overcome numerous obstacles in preclinical and clinical development before achieving regulatory approval. As drug candidates progress, R&D costs increase sharply. In Oncology Drug Discovery, the attrition rate of candidates successfully traversing clinical trials and reaching the market is very high at over 90 percent, and 50 percent of these failures are down to lack of efficacy¹.

 

Reducing these attrition rates underscores the need for improved clinical models, as well as their translation to clinical trial design, analysis and prediction. A significant limitation in developing drug candidates for oncology is the challenge of accessing relevant clinical models that truly mimic the diversity of a patient population and the stages of progression of a disease. As each tumor is unique and cancer progression is non-uniform across all patients, treatments need to be tailored to subtypes of cancer, making discovery and development more complex, time-consuming and costly.

 

 

Over the past decade, treatment of cancer patients has changed considerably to incorporate a more personalized approach. Many patients are now tested for key oncogenic mutations in their tumor DNA, and those who carry a specific mutation may be treated with a targeted therapy. In order to study cancer progression and treatment response with clinically relevant samples, primary tumor cells can be taken directly from the patient and kept in vitro or in vivo to undergo testing. These patient-derived xenograft (PDX) models are capable of providing cost-effective and informative preclinical drug assessments, with large model populations screening drug candidates across a wide range of tumor subtypes, in order to assess the efficacy of a drug candidate before advancing it into the clinic.

 

 Traditionally, cell line- derived xenograft models (CDX), which use cell lines maintained in plastic and therefore adapted to grow independently of the tumor microenvironment, produced genetic and phenotypic characteristics distinct from those seen in the clinic². This led to only 30-40 percent success rates in predicting the clinical efficacy of anti-cancer modalities³. Many aspects of the tumor microenvironment need to be understood and considered, and therefore it is unsurprising that anti-cancer agents developed using simplified models have not yielded the hoped for success in the clinic. In an attempt to improve clinical predictivity and reduce drug attrition, PDX are being used to improve and refine preclinical xenograft modeling, by providing a more relevant heterogeneous system in which human tumor and stromal cells are in close cooperation within a unique environment.

 

 

Establishing models that fully recapitulate the tumor microenvironment both in vitro and in vivo can be challenging. PDX models have been reported to sustain molecular, genetic and histological heterogeneity of the original tumors, and as such, data generated from these models closely resembles clinical data, with over 90 percent prediction of tumor sensitivity and resistance⁴. However one challenge that needs to be overcome is that PDX lose human stroma.

 

Increasing evidence suggests that interactions between stromal and tumor tissues contribute significantly to cancer development and growth, with cancer-associated fibroblasts (CAFs) reported to account for over 50 percent of the tumor mass in some tumors. Stromal reactions to tumors have also been linked to drug resistance e.g. desmoplasia, where the buildup of fibrous tissue protects the tumor from the toxic effects of chemotherapeutics. In order to retain the human stroma in both PDX and cell line models, human mesenchymal stem cells, or patient-derived CAFs, can be supplemented with the tumor cells prior to implantation.

 

 

Retaining primary cancer cells derived from PDX tumors in passage ensures that the models retain more clinically relevant oncological materials than conventional cell lines passaged for decades. Such platforms provide a shortcut to functional screening using the right PDX models with the right drugs, at the right doses, for in-vivo testing, which intimately links together the robustness of the in-vivo analyses with the clinically relevant in-vivo models. In addition, the close relationship to the original patient tumor supports the preservation of genetic features and biological heterogeneity for novel therapeutic target discovery and functional validation.

 

Well-characterized PDX models are used to mimic human clinical trials in mice, known as patient avatar trials or human surrogate trials, providing a clinically relevant diversity of patient population to help identify or confirm responders and non-responders in a population. By identifying suitable biomarkers in response to treatment, Phase 2 clinical trials can be designed to improve the chances of success of a drug candidate in the clinic, thereby reducing the overall cost of drug discovery.

 

 

The use of PDX models in Phase 2-type studies or human surrogate trials is significantly accelerating the pace and reducing the costs of evaluating compounds prior to transitioning into the clinical setting. These models provide a significantly higher level of confidence for decision-making in the drug discovery process and can provide deep biological insights into the pharmacological mechanisms of a drug, helping to identify potential biomarkers important to clinical trial design.

 

The need to reduce drug attrition is especially acute in the field of oncology, where drugs often fail not because of toxicity but rather lack of efficacy. By performing mouse xenograft studies at significantly lower cost and in reduced time, it is possible for drug discovery and development companies to make better decisions faster in their drug development programs before proceeding into the clinic. In addition, the identified biomarkers will help to improve the success rate in development by providing more predictive screening platforms and selection tools to help identify patient populations that will benefit from specific treatment regimens.

 

Preclinical models that more closely replicate the microenvironment and heterogeneity of the human tumor are having a significant impact in guiding clinical strategies and patient selection in the quest to optimize novel cancer therapeutics. The tumor microenvironment has a significant role in disease progression and drug resistance. Understanding the mechanisms of resistance in tumors from any given patient is critical in identifying the most appropriate follow-on therapies as part of a personalized medicine strategy. These models not only provide a greater understanding of the role of the tumor microenvironment and the impact on disease progression, but also offer the opportunity for new drug-target identification and validation.

 

The future of cancer treatments lies in the development of tailored strategies for subsets of cancer patients and by leveraging genomically characterized PDX assets to discover biomarkers and identify patient responder and non-responder profiles. While there is no single treatment that is effective for all patients even within a single cancer type or subtype, by using more clinically relevant models in conjunction with molecular profiling and large-scale data analysis, it is possible to optimize and accelerate drug discovery into the clinic.

 

 

 

¹ Kola I, Landis J. Can the pharmaceutical industry reduce attrition rates? Nat Rev Drug Discov 2004 (3) p. 711-715

² BC Giovanella, JS Stehlin, ME Wall, MC Wani, AW Nicholas, LF Liu, R Silber, M Potmesil. DNA topoisomerase I-targeted chemotherapy of human colon cancer in xenografts. Science 1989, Vol. 246 no. 4933 pp. 1046-1048.

³ Fricker, J. Time for reform in the drug-development process. The Lancet Oncology Volume 9, Issue 12, December 2008, Pages 1125–1126.

⁴ Feibig et al, EJC 40;802, 2004

 

---