For establishment of PDTX models, only small pieces of tumor material are necessary. We aim to implant pieces of 3x3x3 mm in each mouse (min. 1 – max. 4), but smaller biopsies (f.i. fine needle biopsies) can also be used for certain aggressive growing tumor types. However, a sufficiently large piece of tumor material is desirable to preserve part of it for histological and molecular analyses.
Establishment and validation
The time to develop a PDX model depends highly on the tumor type and tumor aggressiveness, and the time to engraft is estimated to take about 2 to 4 months on average, although for some tumor types it may take up to one year. This means that the successful development and validation of a model takes about 1-2 years.
The success rate is the percentage of implanted tumors that effectively engrafted in immunodeficient mice, and that can be propagated successfully to subsequent generations.
In order to validate PDX models, we use different methods comparing the original patient tumor and at least the F3 xenograft tumor. First of all, morphology is checked by a standard hematoxylin and eosin staining and the expression of several tumor specific proteins is examined using immunohistochemistry. Also, the genetic profile is examined using whole-genome low-coverage (copy number alterations), whole exome sequencing and/or SNP analysis, while for some tumor types also transcriptional and metabolomic profiling is performed. Finally, pharmacologic validation or comparing PDX response to a standard of care therapy with the patient response is implemented.
For each xenograft tumor, we store part of the tumor as formalin-fixed and paraffin-embedded tissue blocks, which can be used for (immune)histological analyses. Part of the tumor is also snap-frozen in liquid nitrogen and/or frozen in RNA later solution, in order to be used in molecular analyses. The remaining part of the tumor is preserved as viable frozen tissue, which can be used to transplant in mice at a later time point or to perform cell culture experiments.
Upon request, experiments can be set up to induce treatment resistance in our PDX models. For instance, this can be achieved by treating tumor-bearing mice with a therapy for a prolonged time period until the tumor shows progression under treatment.
The majority of our PDX models are molecularly characterized, which makes it feasible to find targets for personalized medicine strategies. When the presence of a specific target is not yet demonstrated in the available models, analyses can be performed internally or samples can be send out for analyses upon request.
Upon request, new PDX models can be developed, meaning that we can establish more specific models of an available tumor type, but we can also develop PDX models of new tumor types. This is feasible because we have an extensive collaboration with services within the University Hospitals Leuven, enabling us to get access to patient material for the development of our models.
The majority of established PDX models do not metastasize. However, some of our own established PDX models, including melanoma, and ovarian and endometrial carcinomas, have been shown to develop metastases in lymph nodes and/or distant organs.
Although we prefer to implant tumors in the interscapular fat pad, in order to facilitate follow-up of tumor growth and therapy response, orthotopic PDX models can be developed upon request. Experiments are preferably combined with medical imaging for adequate follow-up of these models.
All mice are housed in the Animal Facility of the Katholieke Universiteit Leuven in accordance with legal requirements and ethical codes. The mice are housed in Individually Ventilated Cages (IVC), in which every cage is a barrier and animals are kept free of pathogens. The system contains of an air unit and a rack with cages. The air is filtered and blown in and removed from each cage separately. Environmental enrichment is always available in the cages to allow expression of a range of normal behavior and to avoid physical and social stress.
Mice are housed in standard T2 IVC cages with a ground surface of 435 cm², which allows housing of a maximum of 5 (mouse weight: 25-30 g) or 4 mice (mouse weight: >30 g) per cage.
A humane endpoint can be defined as “the earliest indicator in an animal experiment of (potential) pain and/or distress that, within the context of moral justification and scientific endpoints to be met, can be used to avoid or limit pain and/or distress by taking actions such as humane killing or terminating or alleviating the pain and distress’. This implies that a humane endpoint does not necessarily mean the humane killing of the animal, but it can also result in interventions to alleviate the stressful/painful experimental procedure or providing analgesics.
Humane endpoints that we apply to our PDTX models include, but are not limited to:
- Tumor volume > 1500-2000 mm³
- Tumor causes severe clinical symptoms due to its location, invasive growth or ulceration
- Behavior and locomotion is severely influenced by tumor growth
- General welfare of the animal is severely influenced: circulation and respiration abnormalities, weight loss (15% in 2 days or 20% compared to start of the experiment)
Humane endpoints should always be balanced against the scientific endpoints to be met since pain and distress might be intrinsic to a certain experimental model. However, in this case, the humane endpoint should never be beyond the scientific endpoint.
To allow social interaction, mice are never housed alone in a cage, unless under strict circumstances. Environmental enrichment, such as cage equipment and nesting material allows the animals to interact with and manipulate their environment, and it gives them the opportunity to hide from cage mates, people and unexpected noises, avoiding social and physical stress. For this reason, all cages are enriched with strips of paper fibers, which enables the mice to build nests and hiding places, according to their specific behavior.
At Trace, we are constantly developing alternatives to animal testing. The 3 Rs in animal research are an integral part of our research policy. First of all, we reduce the number of animals used to a minimum, to obtain information from fewer animals or more information from the same number of animals (in vivo medical imaging techniques and/or sequential blood and tumor biopsies). Second, we refine the experiments to make sure animals suffer as little as possible. And third, wherever possible, we replace the use of animals with alternative techniques, or avoid the use of animals altogether. For this reason, we are developing in vitro techniques based on cell lines derived from our PDTX models (PDTX-derived cell lines, PDCs), allowing us to perform a full range of assays on the same tumor entity.
For all our experiments, we are using immunocompromised mouse strains. During establishment of the PDTX models, we use different strains with different levels of immunodeficiency, depending on our previous experiences with tumor engraftment success rates or upon request of research groups. Our most commonly used mouse strains are: scid-beige mice (C.B-Igh-1b/GbmsTac-Prkdcscid-Lystbg N7) with a lack of T and B lymphocytes, defects in cytotoxic T cells and macrophages, and impairment of NK cell functions; CIEA NOG mice (NOD.Cg-Prkdcscid Il2rgtm1Sug/JicTac) lacking mature T, B, and NK cells, with dysfunctional macrophages and dendritic cells and reduced complement activity; and NMRI nude mice (BomTac:NMRI-Foxn1nu) with an abnormal thymus causing a deficiency in T cell function. For treatment experiments, we prefer to use NMRI nude mice, unless otherwise requested.
Our standard anesthesia protocol consists of a combination of ketamine and medetomidine, administered intraperitoneally. After surgery, subcutaneously administered atipamezole is used as an antidote. We also administer analgesia (buprenorphine) as postoperative painkilling. Other anesthesia/analgesia protocols can be considered upon request. Mice are being monitored frequently after any procedure (within a couple of hours after surgery or administration of materials/agents). Monitoring will occur at least once daily or more often if appearing uncomfortable, ill or in pain, and appropriate actions will be taken.