Novel Research Methods Could Expedite New Drug Development

Charles M. Rudin, MD, PhD

Optimizing the methods for preclinical research with an emphasis on patient-derived models, may help speed up the translation of new treatment advances from the laboratory to the clinic, according to a presentation by Charles M. Rudin, MD, PhD, at the 2019 International Lung Cancer Congress.

"Bench to beside is not a one-way street. It is there and back again. It is how can we do discovery research in the laboratory and turn it rapidly into clinical development and patient care but then really, also, in the reverse direction, it's using patient data to try to inform which research questions are relevant," said Rudin, chief, Thoracic Oncology, Memorial Sloan Kettering Cancer Center (MSK). "Optimizing preclinical models could enhance both discovery science and clinical translation."

There are many available models for preclinical research, all of which have limitations for immediate applicability to the clinical setting. The two most common models are cancer cell lines and human cell line xenografts. Both of these options allow for rapid and affordable research; however, there is poor correlation with clinical outcomes. With the growth of genetic manipulation, genetically engineered mouse models have grown in popularity, although these models are restricted by a lack of heterogeneity and complexity.

Two of the more clinically relevant models are patient-derived xenografts (PDX) and circulating tumor cell-derived xenografts, Rudin said. These models have the heterogeneity and complexity that is reflective of human disease; however, they are often not susceptible to genetic engineering and may represent an immunosuppressed microenvironment. Overcoming some of these liabilities could permit for more rapid development, he noted.

To this end, Rudin said that MSK has devoted extensive resources to developing a PDX library for various types of thoracic malignancies, with more than 250 banked samples. These PDX lines were subjected to extensive sequencing using MSK-IMPACT to characterize genetic alterations. As an early example of the success of this approach, Rudin described a study looking at resistance to cisplatin and etoposide in small cell lung cancer (SCLC).

In this preclinical study, molecular characteristics were examined between the mice models with an initial complete response to cisplatin and etoposide and those with a partial response. These findings were then compared with their human counterparts from which the cells were derived, uncovering a strong correlation between the response in mice and those seen in humans.

The next step for the research was to determine why sensitivity to cisplatin and etoposide was lost after 6 months of treatment. From this, the gene SLFN11 was identified as a potential mechanism for resistance. Those with SLFN11 expression were chemosensitive whereas those with SLFN11-low expression were more chemoresistant. Moreover, expression of this gene typically diminished following initial treatment with cisplatin and etoposide.

"One of the genes we think is very important is SLFN11. SLFN11 is thought to be the single strongest determinant of sensitivity to TOP1 inhibitors across all the cell lines, and this is silenced with chemoresistance," said Rudin. "We can show that SLFN11 is epigenetically modulated, and we found that EZH2 inhibitors can turn this gene back on and resensitize tumors."

The ability to induce SLFN11 expression with an EZH2 inhibitor was quickly translated to a clinical study that is currently ongoing at MSK, with the goal of preventing chemoresistance in patients with SCLC. In the phase I/II trial, the oral EZH1/2 inhibitor DS-3201b is being combined with fixed-dose irinotecan for patients with recurrent SCLC (NCT03879798).

"This is just one of many examples that we might bring forward. This was a previously unsuspected pathway in the laboratory that we were able to define on the discovery side and quickly translate because of these representative models," said Rudin.

One of the challenges with PDX is the inability to effectively perform genome editing but Rudin noted that researchers at MSK are looking to overcome this hurdle. For this, a Cas9 marker was introduced into the PDX genes, using an ex vivo spin transduction with a pSpCTRE lentivirus. This allowed for 90% of the PDX tumors to have inducible Cas9 markers, Rudin noted. The approach was verified using EGFR-targeted gene editing, which was capable of producing various sensitivity states for the EGFR gene.

"Now we have models where we have the same EGFR-mutant tumor with sensitivity, with T790M, and C797S, all in the same complex genetic environment," said Rudin. "I think this will be a way that we can accelerate drug development to the clinic."

<<< 20th International Lung Cancer Congress