Patient-derived xenograft (PDX) models lead the way in targeted cancer therapy
Health & Medicine
ArticleDetailDownload PDF
Researchers at Vanderbilt University Medical School are using in vivo murine patient-derived xenograft (PDX) models in pre-clinical trials of acute myeloid leukaemia (AML). These models – in which tumorous tissues are transplanted into immunocompromised mice – serve as powerful tools for testing investigational new drugs for the treatment of AML. Lead researcher Dr Haley Ramsey is utilising PDX models to help clinicians get the right drugs to the right patients.
Globocan (an online database that provides global cancer statistics) reported an estimated 19.3 million new cases of cancer and close to 10 million deaths as a result of cancer worldwide in 2020. Although the scientific community has made great strides in understanding the pathogenesis of many types of cancer and developing therapies, it remains a complex disease that continues to evolve and grow amongst the human population.
When we look at cancer, we see more and more that the disease can be specific to the individual, with biomarkers and hallmarks that are not common in all patients suffering from the same type of cancer. Over recent years, biotech companies have begun to research customised cancer therapies to target specific characteristics of a patient’s cancer, with the aim of developing a more tailored therapy for the individual. In the past, a majority of cancer research has been completed in immortalised cell lines. While helpful in uncovering cancer mechanisms, this approach leaves little to be learned in the way of cancer’s heterogenicity. Because primary cell cultures derived from cancer patients typically only survive for a few days, methods were needed for the carried growth expansion and support of unique patient samples. Such a need led to the first models implanting human tumour tissue into inbred mice over fifty years ago (Rygaard & Poulsen, 1969).
unoL/Shutterstock.com
The power of PDX
The term patient-derived xenograft (PDX) describes an in vivo model whereby tissues or cells from a cancer patient’s tumour are transplanted into a mouse. Such recipient mice are termed humanised mice, meaning they have been engineered to express human genes in an attempt to greater mimic human environment necessary to propagate human cells. These mice models are as close as possible to human physiological conditions and are used to study cancer in vivo. They are an integral part of pre-clinical medical research.
The beauty of PDX transplants is that researchers can study patient-specific cancers using this model. The natural growth of the cancer, its characteristics, changes in mutations, and its response to treatments can all be studied using PDX, making this a powerful treatment tool. At Vanderbilt University Medical School, Dr Haley Ramsey has created one of the largest collections of fully annotated PDX mouse models in the world. The collection of over 30 unique PDX models are used for pre-clinical testing of novel therapies in acute myeloid leukaemia (AML).
“The use of murine PDX models has greatly strengthened the translational value of pre-clinical medical research.”
Drug resistance in AML
AML is a rare form of blood cancer that is characterised by the growth of abnormal white blood cells (myeloblasts) in the bone marrow (BM). As the cancer progresses, the BM fills with AML myeloblasts, with eventual growth into the bloodstream. The patient samples for the PDX trials are collected from leukapheresis, a laboratory procedure that removes white blood cells or – in the case of AML – cancerous blast cells from whole blood.
Despite intensive research efforts, few therapeutic advancements in AML have been made in the past 40 years and mortality remains high, with the overwhelming majority of patients succumbing to the disease within five years of treatment. Dr Ramsey uses her expertise within a translational research team that has made efforts to maximise the ex vivo study of human primary AML cells with benchtop experiments – reserving the most refined, final questions to be tested via in vivo murine work. In the end, the researchers’ ability to develop biomarker-driven AML therapies will rely on the validation of findings in the best avatar we have for human beings. Currently, that is the PDX transplantation.
Venetoclax is currently one of the most common therapeutics employed in AML treatment, which specifically targets the BCL-2 protein in cancerous cells. BCL-2 is overexpressed in some AML patients and it prevents the cancer cells from dying, resulting in continued growth of the cancer. Venetoclax blocks the BCL-2 protein and restores apoptosis and thereby causes cancerous cells to die. Venetoclax was recently approved by the FDA for use in elderly AML populations, transforming the standard of care for this disease. Still many do not respond to initial therapy and relapse is common. Moreover, in cell lines venetoclax treatment can lead to upregulation of other anti-apoptotic proteins, giving rise to venetoclax resistance. The researcher’s previous work suggests that resistance to BCL-2 inhibition is hallmarked by dependence on MCL-1 and increased sensitivity to MCL-1 inhibition.
MCL-1 is an anti-apoptotic protein and member of the BCL-2 protein family, which is commonly upregulated in AML. In a 2018 paper, Dr Ramsey and her colleagues describe the potential use of a potent small-molecule inhibitor of MCL-1 (VU661013) for AML. Initially, the team tested VU661013 activity in AML cell lines in high-throughput screening and found growth inhibition of most of the cell lines tested. These initial observations prompted the researchers to test VU661013 in one of their cell-line derived xenograft models (CDX) with the MV-4-11 human AML cell line. Cell line models are often used first to test for initial efficacy before moving onto precious patient sample derived PDXs.
Graph depicting the weekly measurement of peripheral blood cancer of a PDX model treated with venetoclax and a test compound (left). The pictures on the right depict bone marrow undergoing the same treatment. The bone marrow cells are stained brown from an antibody that marks human cancer cells, while normal healthy mouse cells stain blue.
The results showed that mice treated with VU661013 incurred significant decreases in tumour burden. Further studies revealed that combination therapy of venetoclax and VU661013 in the same CDX MV-4-11 AML cell line and an additional MOLM-13 AML cell line CDX model resulted not only in decreased tumour burden in the blood, bone marrow and spleen but also in a significant prolongation of lifespan in diseased mice.
“Our power is in the full clinical and genetic annotation we have in these models.”
Bone marrow samples from three patients that had previously failed venetoclax treatment (relapsed or refractory) showed effectivity when treated with the combination of VU661013/venetoclax. These experiments showed for the first time that MCL-1 inhibition can overcome venetoclax resistance. The use of patients’ own cells in a mouse model means physicians are able to ascertain certain hallmarks of the cancer and tailor the best therapy or combinations accordingly in future studies and clinical trials.
Following the synergy noted between venetoclax and VU661013, the team now turned again to a PDX model to test the dual inhibition of MCL-1 and BCL-2. Data showed that one of the three BCL-2 dependent patients did not benefit much from the addition of VU661013 to venetoclax. However, in the MCL-1 dependent patient, combination treatment decreased tumour burden in comparison to venetoclax treatment alone. Again, PDX proved a powerful tool here, and the findings suggest that some AML patients can benefit from combined treatment of MCL-1 inhibition and venetoclax in reducing tumour burden.
BET inhibition enhances venetoclax activity in AML
Dr Ramsey continued to investigate other potential drugs that could be used in the treatment of venetoclax-resistant AML, and PDX models were again integral to these studies. Dr Ramsey explains that “the use of murine PDX models has greatly strengthened the translational value of pre-clinical medical research”.
MYC expression is dysregulated in many AML. BRD4 is a BET (bromodomain and extraterminal) family protein that regulates the transcription of MYC target genes. Disruption of BRD4 at the MYC locus leads to the silencing of MYC gene expression and reduction in cell proliferation. This makes BET family proteins a good target for AML therapy.
Dr Ramsey led inquiry into a novel BET inhibitor (BETi), INCB054329, for its potential to promote apoptosis of leukemic cells in combination with venetoclax. After initially testing the combination in AML lines, results supported the coadministration of BETi and venetoclax in inhibiting transcription of key oncogenes and reduced cell cycling. Moreover, this combination was synergistic in both venetoclax-sensitive and venetoclax-resistant cell lines.
Leukapheresis is a laboratory procedure that removes white blood cells or cancerous blast cells from whole blood.
The team moved the experiments into one of their AML PDX mouse models. Results showed that BETi treatment stopped the cell cycle and reduced BCL-2 levels, allowing the combination of venetoclax to be administered at a lower dose. This combination treatment also reduced venetoclax-associated marrow toxicity. More importantly, the researchers observed that this therapy-induced apoptosis in AML hematopoietic CD34+ stem cells but not in normal haematopoietic CD34+ stem cells, meaning normal cells are unaffected. These findings were a huge step forward in BETi studies, promising safer combination therapy for AML.
Proof-of-concept for oral anti-cancer drug delivery
DNA methyltransferase inhibitors (DNMTis) are used in the treatment of myelodysplastic syndrome and AML. DNMTis used include azacitidine (AZA), administered subcutaneously (s.c.) or intravenously (i.v.) daily for 5-7 days. The demands on patients for parenteral dosing of AZA are not trivial. Commonly, patients experience pain associated with multiple intravenous punctures, associated risks of infection, bleeding and thrombosis, and inconveniences associated with central catheters or s.c. infusion ports. Moreover, when oral vs. intravenous treatment strategies have been assessed in other malignancies, patients strongly preferred oral agents. To this end, there has been significant interest in the development of an oral formulation of DNMTi, but attempts to provide the therapy orally have been limited given rapid clearance of the agents by the enzyme cytidine deaminase. Cytidine deaminase (CDA) is an enzyme found in the gut and liver which rapidly clears DNMTis when ingested, making it difficult to provide this therapy orally. An oral inhibitor of CDA – cedazuridine (CDZ) – shows some promise in inhibiting the action of CDA on DNMTis, and thus allows DNMTis to be effective when taken orally. Dr Ramsey and colleagues decided to evaluate if oral dosing of AZA could be achieved using CDZ, and they again utilised their PDX models to test this.
When oral AZA and CDZ were administered in human AML cells, they exhibited a pharmacokinetic profile and efficacy similar to AZA administered either s.c. or i.v.. Moreover, the oral AZA and CDZ treatment caused tumour regression in a MOLM-13 CDX. Next, the lab combined this oral regimen with venetoclax in a PDX model and found that the response to this treatment was comparable to that seen with venetoclax and AZA given in their traditional forms (i.e., AZA given by i.v. or s.c.). These studies again demonstrated the usefulness of PDX models, this time to provide proof-of-concept for oral DNMTi administration for the treatment of myeloid disorders.
Dr Ramsey shares, “Our power is in the full clinical and genetic annotation we have in these models: we know the genotype of the patient, previous treatments, and have full clinical flow cytometric analysis from AML diagnosis”. Utilising leukapheresis samples for PDX models has helped the team make great progress in the field of AML therapy over the last half a decade.
What are your next research aims?
While increases in BCL-2 production are most likely behind a majority of venetoclax resistance cases, I aim to define the other etiologies behind the resistance. This means we will begin to look at the behaviour of the cancer cell, how it respires and fuels itself – a view far beyond just its detectable mutations. Such information may be able to allow us to find new biomarkers to predict where the cancer is headed in a post-venetoclax scenario. To know the impending mechanism of venetoclax resistance will become the first steps in being able to strategically overcome that situation.
Ramsey, H. E., Fischer, M. A., Lee, T., Gorska, A. E., Arrate, M. P., Fuller, L., Boyd, K. L., Strickland, S. A., Sensintaffar, J., Hogdal, L. J., Ayers, G. D., Olejniczak, E. T., Fesik, S. W., & Savona, M. R. (2018). A Novel MCL1 Inhibitor Combined with Venetoclax Rescues Venetoclax-Resistant Acute Myelogenous Leukemia. Cancer discovery, 8(12), 1566–1581. Available at: https://doi.org/10.1158/2159-8290.CD-18-0140
Ramsey, H. E., Greenwood, D., Zhang, S., Childress, M., Arrate, M. P., Gorska, A. E., Fuller, L., Zhao, Y., Stengel, K., Fischer, M. A., Stubbs, M. C., Liu, P., Boyd, K., Rathmell, J. C., Hiebert, S. W., & Savona, M. R. (2021). BET Inhibition Enhances the Antileukemic Activity of Low-dose Venetoclax in Acute Myeloid Leukemia. Clinical cancer research: an official journal of the American Association for Cancer Research, 27(2), 598–607. Available at: https://doi.org/10.1158/1078-0432.CCR-20-1346
Ramsey, H. E., Oganesian, A., Gorska, A. E., Fuller, L., Arrate, M., Boyd, K., Keer, H., Azab, M., & Savona, M. R. (2020). Oral Azacitidine and Cedazuridine Approximate Parenteral Azacitidine Efficacy in Murine Model. Targeted oncology, 15(2), 231–240. Available at: https://doi.org/10.1007/s11523-020-00709-x
Rygaard, J. & Poulsen, C.O. (1969). Heterotransplantation of a human malignant tumour to “nude” mice. Acta Pathologica Microbiologica Scandinavica, 77, 758-760. Available at: https://doi.org/10.1111/j.1699-0463.1969.tb04520.x
Dr Haley Ramsey uses PDX models in pre-clinical studies for novel therapies for acute myeloid leukaemia.
Collaborators
Dr Stephen Fesik (Vanderbilt University)
Dr Brian Bachmann (Vanderbilt University)
Dr Scott Hiebert (Vanderbilt University)
Bio
Haley Ramsey received her BSc from the University of Tennessee, an MSc from the Science Academy of Bonn-Rhein-Sieg and her PhD from the Medical University of Vienna, where she was credited with creating the first murine model of alloreactive mismatch T memory cells for use in bone marrow, heart, and skin transplantation studies. Prior to joining the Savona Lab in 2015, she was a postdoctoral fellow at Harvard Medical School. She currently designs pre-clinical studies for novel therapies in Acute Myeloid Leukaemia and the effects of therapy on tumour metabolism.
Haley Ramsey
Contact
Haley Ramsey, PhD
Research Assistant Professor
Vanderbilt-Ingram Cancer Center
2220 Pierce Avenue
Preston Research Building 518
Nashville, TN 37232
USA
