Table 1

Criteria for testing a therapeutic treatment in NOD mice

• Ensure NOD females have a high incidence of diabetes (e.g., ≥70% diabetes incidence by 300 days of age) in the chosen animal facility if a preventative study is being performed
• Establish an SOP and fixed study entry criteria for the treatment regimen
• Perform a power calculation to determine the minimum cohort size required to detect an effect of the desired size
• Determine what controls are required to evaluate treatment efficacy
• Monitor diabetes incidence in controls contemporaneously with treated mice
• Ensure cohorts are tested in a timely fashion to avoid variable diabetes incidences over a long period
• Use pharmaceutical-grade reagents when possible
• Generate sufficient reagent stocks for treatment of all mice to minimize effects due to different reagent preparations
• Randomize treatment within age-matched littermates
• Have the same experienced personnel perform the treatment and monitoring protocols for all mice
  • Ideally, these criteria should be identical for single-lab and multicenter trials. However, it is well recognized that initial testing of a proposed treatment by a single lab may not be optimal because of various factors (e.g., the most effective treatment protocol is not obvious, pharmaceutical-grade reagents are not readily available, or technical proficiency for performing treatment and monitoring protocols may be limited). This should not preclude the publishing of such studies as long as the caveats for the results are clearly described by the investigators. Indeed, the advantage of the NOD mouse model is that a promising treatment protocol can be further tested and potentially improved upon by other groups.