Patient-derived xenografts (PDX) are arguably the most faithful and the most powerful modeling system in oncology today. And while PDX models are far from being perfect, most researchers find them extremely useful in studying cancer biology, biomarker development, drug discovery, and drug response. To date, no other system can provide a more accurate representation of human cancer than a PDX model can.
Understanding Patient-Derived Xenografts (PDX)
Basically, PDX is a representative model of cancer, where cells or tissues harvested at biopsy or surgery from a patient’s tumor are directly implanted into immunodeficient mice for research purposes. Using immunodeficient mice (i.e., athymic nude, severely compromised immune deficient or SCID, non-obese diabetic SCID or NOD-SCID, and Rag2-knockout mice) prevents transplant rejection and ensures successful xenotransplantation in the mucine host.
In PDX model studies, necrotic tissues are removed from the resected tumor before the tumor is physically or mechanically sectioned into discrete fragments or single-cell suspensions. The tumor fragments are then implanted heterotopically (implanted subcutaneously or in subrenal capsular sites that are unrelated to the original tumor site) or orthotopically (implanted into the corresponding anatomical position) into a mucine host.
Generally, 5 to 10 mice are used to establish the first generation (F0) of PDXs. Implanting the initial tissue into several mucine hosts increases the chance that there will be enough cells for further experiments. The tumor is passaged over to the next generation of mice (hereafter denoted as F1, F2 … Fn) when the tumor-burden becomes too large for the F0 mice. Tumor engraftment may take about two to four months while engraftment failure may be declared if there is no palpable tumor after six months.
PDX models can be used to test to see the effect of certain types of treatment before being administered to the patient, create individualized treatment plans, and in the development of new cancer drugs.
As mentioned earlier, the PDX model has its own limitations. Since it uses immunodeficient mice, it cannot be used in studying anti-cancer agents and immunotherapies targeting the immune system components. To address this issue, researchers are starting to explore the potentials of using humanized-xenograft models which can be created by co-engrafting the patient tumor fragment and bone marrow or peripheral blood cells into a NOD-SCID mouse.
However, further studies are needed to validate if it will work on all tumor types and determine if the reconstituted immune system will function in the same way as the patient’s.
Cancer Cell Lines vs. PDX Models
Unlike cancer cell lines, which were derived from tumor cells and proliferated within in vitro cell cultures, PDX tumor models were not grown in plastic. Instead, they were propagated in micro-environments that closely resemble that of the patient’s. As a result, they produce heterogeneous tumors that do not exhibit significant genetic alteration, despite being passed on from one generation to the next.
On the other hand, cancer cell lines produce phenotypically homogeneous cultures that do not easily develop tumors and are prone to producing genetically divergent tumors when implanted into immunodeficient mice.
This is quite interesting, since according to a study published in the March 2011 issue of Nature Reviews Clinical Oncology, the lack of tumor heterogeneity is one of the main reasons for the high drug attrition rates for cancer. In fact, only 5% of anti-cancer agents are licensed by the FDA after exhibiting sufficient efficacy in phase III testing.
Moreover, since implanted tumor tissues in PDX models retain all the patient’s genetic and epigenetic abnormalities, they are often more predictive of the patient tumor response to certain anti-cancer agents.
Using PDX models, oncologists were able to make significant headway in understanding breast, colorectal, pancreatic, and pediatric cancer or neuroblastoma. Human-xenograft models have likewise been created for acute myeloid and acute lymphoblastic leukemia.
