Designing an in-vivo study
The development of new, effective treatments for patients begins with robust studies of the pharmacokinetics (PK – the movement of a treatment through the body) and pharmacodynamics (PD – the response of the body to the treatment). PK studies look at absorption, distribution, metabolism, and excretion of the treatment while PD studies link the concentration of the treatment to its safety and efficacy. The Animal Imaging and Preclinical Testing Core Facility at the KI provides the expertise and equipment to design and perform the essential PK/PD experiments needed to analyze the therapeutic potential of your novel treatment. We are dedicated to helping researchers who have developed a product that needs to be tested in an animal model, but who lack either the resources or skills needed to perform that testing.
Nearly all of the experiments described here are pilot studies. The purpose of a pilot study is to provide information about the feasibility and parameters necessary to design a larger preclinical study that will produce statically robust data. A pilot study will not produce statistically significant results, but it does provide some information on variability. The data from a pilot study is subjected to a power analysis to determine the sample size that is needed to provide statistical power to detect a clinically relevant difference between a control and treatment group for the preclinical study. Data generated from a pilot study should not be included in the analysis of the larger preclinical study.
Click here for Michael FW Festing’s site or here for the National Centre for the Replaccement, Refinement & Reduction of Animals in Research site. Both are extensive websites dedicated to experimental design for research scientists working with laboratory animals, with the dual goals of reducing the numbers of animals used in research while producing high-quality, reproducible data.
Please read MIT’s Committee on Animal Care Policy on Minimizing the Number of Animals Used to Obtain Valid Results (Requires MIT Certificates)
Here is an overview of the steps that lead up to and include robust preclinical testing of a new treatment for efficacy. Not all steps apply for every project.
Perform in vitro toxicity studies (not required, but very useful)
Determine the concentration that is lethal to 50% of cultured tumor cells (LC50).
Determine lethal concentration (LC50) for primary or cultured non-tumor rodent cells
Choose your animal model Please meet with the Preclinical Projects Manager, Aurora Connor to discuss your options.
- If you will be injecting a human tumor cell line into rodents, you must use immune-compromised animals. There are many different choices, including Nude, SCID, NSG, NRG and more. A great table of the properties of the 8 most commonly used mouse strains is available here. Only Nude rats are commonly available, though SCID rats are being developed by several vendors.
- If you will be injecting mouse tumor cells into mice, the sub-strain of the mouse that produced the original tumor should be used for your study to prevent an immune response. If you will eventually be using a known mouse model of disease, pilot studies should be performed in the same substrain of mouse, though it does not need to be the disease model for those studies. For example, the most commonly used strain laboratory mouse is called Black6 or C57BL/6. However, it has several common sub-strains with different single nucleotide polymorphism (SNP) profiles including C57BL/6J, C57BL/6JJcl, C57BL/6JOlaHsd, C57BL/6NJcl, C57BL/6NTac, and C57BL/6CrSlc. These polymorphisms include alterations to genes and results in phenotypes related to diet induced diabetes, immune response, Parkinson’s disease, retinal development and others. Using the same substrain throughout your project is critical to getting useful and reproducible data. For more information, see this publication and a Jackson Laboratories webinar on this subject.
- If you will be using bioluminescent or fluorescent imaging (with the IVIS machine) and you have a wide choice of mouse strains, using a white mouse such as Balb/c will produce better results because they do not have pigment in their skin. Furless (nude) mice will produce the best imaging results, but these mice are immunocompromised and may not be ideal for all studies. Mice should be fed alfalfa-free pellets for 1 week prior to fluorescent imaging to reduce the levels of background produced by chlorophyll in the stomach and intestines.
- There are many genetically engineered strains of mice that will develop cancers and other diseases in a manner that closely models human disease progression. Choosing the model that most accurately mimics the disease state you are interested in can be tricky. Some of these mice are already housed at the KI, many others are available commercially from distributors like the Jackson Laboratories, and Aurora is happy to help you with this decision. Again, these mice are on many different substrain backgrounds and it is important to be consistent throughout your project.
Determine Maximum Tolerated Acute Dose (MTD) in rodents
- Initial MTD studies are usually performed in mice, even when the eventual desired animal is a rat, because it is more cost effective and uses up far less of your compound. Dosing in mice can be effectively determined for rats, so this is a good initial step.
- For injection, the compound must be soluble in an aqueous solution and should be pH 7.0-7.5 if possible. For oral dosing, solubility and pH requirements are less stringent.
- If your compound is hydrophobic or otherwise insoluble, addition of a caging molecule such as a cyclodextrin can help with solubility. An extensive list of common vehicles used in animal models for various dosing routes can be found here.
- Generate a hypothesis of a likely dosage based on in vitro data and known dosages of similar compounds. This paper used over 300 chemicals for whom the in vitro LC50 (concentration at which 50% of cells die) and the rodent LD50 (concentration at which 50% of animals must be euthanized) were known. They found that in general, to convert from in vitro toxicity (LC50) to in vivo toxicity (LD50), the equation is:
log(1/LD50)= -1.1 + 0.36xlog(1/LC50)
This equation does not hold true for nanoparticles or antibodies.
- Based on a best guess dosage of y, perform an initial MTD study in mice:
There should be 4 cohorts of 2 mice each (one male and one female):
Cohort 1- vehicle
Cohort 2 – 1/10y or 1/5y
Cohort 3 – y
Cohort 4 – 5y or 10y
If no or minimal toxicity is seen, a new study is performed with 50y, 250y and 1250y (or 100y, 1000y, 10,000y if desired)
If the lowest dose is toxic, the study is performed again at lower doses.
When you have found a toxic dose reduce the concentration, narrow your dosing window and triple the number of mice per cohort:
For a mixed-gender study, you need 3 cages of 4 males and 3 cages of 4 females. Each cage should house one mouse from each cohort, 24 mice total.
If the cell type is gender specific such as breast or prostate and the study has to be single sex, you need 3 cages of 4 mice. Again, each cage should house one mouse of each cohort. Ear punching is an easy way to identify each cohort. Another tagging option is Sharpie or tattoo on the base of the tail. Sharpie marks wear off after 2-3 days, though.
Mice are weighed and dosed on Monday, then weighed each subsequent day. If a mouse loses more than 10% of its original bodyweight or appears to be suffering, it should be euthanized and necropsy should be performed. If only minimal weight loss and no adverse behavior is observed, necropsy with blood collection occurs one week after dosing. Blood is collected for whole blood counts as well as liver and kidney enzymes. Liver, kidney and sometimes lung, stomach or other organs are collected depending on the study.
Determine the clearance rate and biodistribution of the compound
- Clearance of the treatment is most easily performed by dosing mice with the Maximum Tolerated Acute Dose (MTD) then taking blood samples at regular time points over the next 24-48 hours. The blood is processed by LC/MS to determine when the treatment has effectively cleared from the circulating bloodstream. For treatments that are supplied orally or through a non-intravenous injection, this analysis also provides information about when the maximal concentration is reached and how long it persists after dosing.
- Some treatments can be labeled with a fluorescent dye, can be specific to the expression of bioluminescence, or linked to a radioactive label for visual and quantitative biodistribution using our IVIS and PET imaging systems, respectively.
- After you have determined when the treatment is cleared from the blood, you can also perform a study where you euthanize mice at a critical timepoint, harvest and homogenize the organs, and run these samples through LC/MS to look for relative distribution to your tissue (organ or tumor) of interest.
Determine the Maximum Tolerated Dose with multiple doses
- Based on the acute toxicity and clearance data you learned in steps 3 and 4, perform a repeat dosing experiment to determine the maximum tolerated dose if it is given multiple times over 4 weeks or longer depending on the current standard of treatment in patients. Some treatments can be dosed daily, while others are given once, twice or three times per week.
- Mice must be weighed regularly and at the end of 28+ days, blood and tissues are harvested for analysis by a pathologist.
Other Pilot Studies that May Need to be Completed
- Rate of tumor/disease development or progression in your mouse strain
- Dose optimization of FDA-approved drug (current standard of care) if this has not been published for your strain of mouse.
- Surgeries to optimize placement of a device.
Perform a Pilot Study for Efficacy of your Treatment
- Similar to the design of the MTD studies, There are often 4 cohorts of mice each (4-5 of each sex):
Cohort 1 – vehicle (saline, 10% cyclodextrin for example)
Cohort 2 – the current treatment compound (cisplatin for example)
Cohort 3 – your compound at MTD
Cohort 4 – your compound + current drug
Run the Data from Step 7 through a Power Analysis
The power of an experiment is the probability that the data will show a treatment effect, if it is present. A power analysis is used to estimate the sample size needed for a powerful experiment. If too few animals are used, an important result can be missed, but the use of too many animals is unethical. Both scenarios waste animals, time, and resources.
Please read MIT’s Committee on Animal Care Policy on Minimizing the Number of Animals Used to Obtain Valid Results (Requires MIT Certificates)
Click on this link from Taconic to learn more about Experimental Power and Reproducibility.
A free program to run power analyses is G*Power. It includes calculations for the t-test, F-test (one-way analysis of variance) and others. It can be downloaded from this web site.
Perform a Robust Preclinical Efficacy Study and Present Your Data Well
It seems there are endless options for graphs and plots to present data from preclinical tests. Some make it easy to get your point across quickly for a slide presentation while others are better at representing the most unbiased results of your study. In particular, pharmaceutical companies and manuscript reviewers are increasingly asking for spaghetti plots to show the effect of new chemotherapies on tumor growth. Here are papers from Roche Pharmaceuticals and from JMP Life Sciences on the topic.
As you present your data, your choice of words to describe the animal’s or the tumor’s response to treatment should be the same as those a doctor or clinical trials manager uses. RECIST (Response Evaluation Criteria In Solid Tumors) is a set of published rules that define (among other things) when cancer patients undergoing treatment
- Improve/Respond Completely
- Improve/Respond Partially
- Stay the same/ are Stable
- Worsen/Progress
possible link to study design tool