Keeping expanded access programs in context

In this issue of the Journal, Gelevski and colleagues describe their experience treating 17 ALS patients with IC14, an anti-CD14 mono-clonal antibody, via an intermediate size EAP. 8 Their report is an excellent example of the kinds of helpful data that can be collected from an EAP and showcases some best practices for reporting. The duration of disease at baseline for each patient is omitted but there is enough clinical information presented to hint at a bimodal population, as discussed earlier. At least seven participants have either established slow progression or have clinical patterns that predict slow progression. Another seven participants have either low vital capacities or bulbar dysfunction, suggesting later stage disease. The wide range of treatment duration (1 to 22 + months) and broad distribution of ALS Functional Rating Scale scores (14 to 44) at baseline would not be expected in a clinical trial. The authors report pharmacokinetic and pharmacodynamic data indicating target engagement, safety information, and demonstrate a lack of persistent anti-drug antibodies, all of which would guide decision-making regarding future clinical trials. The study collected ALS Functional Rating Scale progression data before treatment but no formal analysis of pre- versus posttreatment slope was performed, which should discourage any remarks about efficacy. Also very appropriate was a statement highlighting the drawbacks of an EAP design. The increasing availability of EAPs should allow more ALS patients to access potentially efficacious drugs and contribute to the altruistic goal of defeating the disease for all patients. Publication of EAP results should include enough detail about the study population so that readers can place the data in context with “ typical ” clinical trial populations. In addition to measures of safety, pharmacological prop-erties, and target engagement, articles should include baseline vital capacity and duration of disease. If there is any discussion of apparent efficacy, authors should take care to adjust their historical controls, and statistical analysis may need to account for a bimodal pattern of progression. The data from EAPs may be invaluable but the differ-ences between clinical trial and EAP design need to be kept in mind.