In September of 2011, RSRT met with the National Institute of Neurological Disorders (NINDS) and other public and private organizations that fund Rett Syndrome research to discuss crucial knowledge gaps in the field. The main findings of the workshop were published recently in Disease Models & Mechanisms.
In particular, the meeting focused on how the research community can improve its chances of success in clinical trials. Preclinical studies require a huge investment of time and effort studying disease in rodent models. Even then, for a variety of reasons, drugs that show promise in preclinical studies will often fail in the clinic.
Here are a few big hurdles in preclinical animal studies — some that are specific to Rett research — and how experts are meeting those challenges.
1. Studying female mouse models of Rett.
“It’s important that when we do a drug trial, that we really impact features that are clinically meaningful, features that are going to impact patients,” Huda Zoghbi of the Baylor College of Medicine in Houston, Texas, told RSRT in a recent interview.
Like girls with Rett, however, female mouse models of the disorder vary in the type and severity of their symptoms, which makes them harder to study than males.
That’s because the gene missing or mutated in Rett, MECP2, is located on the X chromosome. Female mice — which, like girls, have two X chromosomes, only one of which is active — will have either mutated protein or normal protein levels, depending on which copy is expressed in the cell. Rarely are they missing all of their MeCP2 protein.
In contrast, male mouse models missing the Rett gene have no protein at all. Although these mice have paved the way in understanding the protein’s role in the brain, when it comes to treating Rett, results from studies of male mouse models may not be the ideal model to work with.
More researchers are turning to female mouse models. Zoghbi and Rodney Samaco, also at Baylor, for example, published a study in October in Human Molecular Genetics, describing two different female mouse models of Rett in detail. Detailed characterization of these mice will help lay the groundwork for preclinical studies.
2. Unknowns about how an animal’s environment affects therapeutic efficacy.
No two research labs are alike. The ways in which they differ, including animals’ access to food, housing, lighting or other environmental factors, might well influence an animal’s response to a drug.
What’s more, an individual mouse’s genetic environment — meaning the genetic background on which Rett mutations are made — affects some of its symptoms, such as obesity and abnormalities in social behavior. These genetic differences may also affect how animals respond to treatment.
Variability in genetic and environmental conditions plague scientists studying many conditions, not just Rett syndrome. One way to help address this obstacle, according to Rett researchers in the Disease Models & Mechanisms workshop summary, is to study symptoms and potential therapies across a variety of models and in many lab settings. Those mouse models that show consistent results across different environments will be most useful for translational studies.
3. Recapitulating speech problems in mice.
Of the many symptoms seen in Rett, loss of speech is among the most challenging to study in a mouse model. Some groups have shown that Rett mouse pups produce unusual vocalizations when they’re separated from their mothers in early postnatal life. These sounds are either more or less frequent than in healthy controls, depending on the mouse model studied. Future work will need to sort out these conflicting results and identify a mouse model that best captures this hallmark symptom of Rett, researchers say.
Unintended biases can creep into animal studies. This can lead researchers to conclude a treatment is effective when it isn’t, or it can cause overestimations of a drug’s efficacy.
In recent years, researchers across numerous fields have stepped up efforts to improve study rigor. In June of this year, NINDS convened a panel of scientists, funders and journal editors to talk about how researchers can do a better job reporting methods in preclinical animal studies; both in grant applications and journal publications. At the very least, the panel concluded in a perspective published October in Nature, researchers should report on the following practices:
- Randomization, where animals are randomly assigned to receive either treatment or placebo;
- Blinding, where researchers doing the experiments or analyzing the data are unaware of whether of which animals are receiving treatment or placebo;
- Sample-size estimation, a calculation of an appropriate sample size at the study’s outset;
- And how data is handled, for example, deciding on study’s primary endpoints, or how to handle missing data points or outliers, before starting the study.
Not reporting such details has, in the past, been linked to overestimations of therapeutic efficacy, according to the NINDS report.
Now Rett researchers have added their voices to the mix in the Disease Models & Mechanisms report, voicing their support of NINDS’s recommendations and emphasizing the need for rigorous experimental design.