New Eligibility Criteria Needed for ALS Clinical Trials

January 31, 2019

Approximately 60% of the patients with ALS are already excluded from clinical trial participation at the day of diagnosis, but investigators think that 60% is an underestimate.

Ruben Van Eijk, MD, MSc

The results of a recent literature search concluded that on average, 59.8% of patients with amyotrophic lateral sclerosis (ALS) are excluded from trial participation, which questions the generalizability of trial results.

The most common reasons for trial exclusion was that patients did not meet a specific El Escorial category (23% excluded) followed by respiratory function as determined by forced vital capacity (17% excluded), and required disease duration (12% excluded). Researchers reviewed current practices of participant selection in clinical trials and assessed the effects on trial populations, efficacy endpoints, generalizability, and outcome heterogeneity in ALS clinical trials to measure the value of risk-based inclusion criteria. Investigators performed a literature search to summarize the eligibility of criteria of clinical trials which were applied to an incidence cohort of 2904 consecutive patients with ALS to quantify the effects on generalizability and outcome heterogeneity.

“Approximately 60% of the patients with ALS are already excluded from clinical trial participation at the day of diagnosis,” Ruben P.A. van Eijk, MD, MSc, clinical epidemiologist, University Medical Center Utrecht, told NeurologyLive. “Considering that patients are usually not directly enrolled into clinical trials after diagnosis (e.g. patients need some time to consider, mean time between diagnosis and enrollment is 6 months), the 60% is an underestimate. Probably, only 10%­—20% of the patients are eligible to participate in trials. The safety and efficacy (i.e. do patients benefit?) of medicines is therefore unknown in a large proportion of the patients with ALS.”

The second important finding, van Eijk explained, is that despite the exclusion from clinical trials, there is almost no benefit for the trial itself, "To illustrate, investigators want to exclude patients that do not change during the trial’s follow-up or that are unlikely to complete the trial. We found, however, that these “unfavorable” patients remain in the trial and that the current exclusion criteria cannot identify these patients.”

To harmonize the comparison between clinical trials, investigators included only randomized, placebo-controlled, clinical trials evaluating the efficacy of a single pharmacologic agent and excluded clinical trials investigating multiple agents, solely aiming to determine safety, having non-clinical primary endpoints, or starting enrollment prior to the approval of riluzole. The primary goal for the investigators was to estimate, per the eligibility criteria, the number of patients from the general population that could be considered ineligible to participate.

Investigators included 38 randomized, placebo-controlled clinical trials and concluded that, on average, 59.8% (95% Ci, 52.6%­—66.7%) of patients are found to be ineligible to participate in clinical trials on the day of diagnosis in the incident cohort—exclusion rates varied between trials and ranged from 14% to 95%. The investigators believe that 59.8% is an underestimation of patients excluded from participation since most patients are enrolled in clinical trials a few months after diagnosis and at the time, 15% to 24% of the patients who could be prescribed the drug at diagnosis are deceased and never evaluated. Additionally, a larger proportion of remaining patients fail the criteria due to disease progression, which may lead to exclusion rates in real-world settings that are much larger than those reported in other fields.

“The most important conclusion could be that, for the majority of patients with ALS, we have no information whether drugs evaluated in previous trials work or are safe,” van Eijk added. “This is especially true for patients with a low lung function or long symptom duration, as these patient groups were almost always excluded from trial participation. Moreover, some of these exclusion criteria are indirectly related to other patient characteristics. For example, patients with dementia or cognitive disorders are often excluded, but these factors are also related to a specific ALS-related genetic mutation (i.e. C9orf72). In another study, we showed that the working of a drug might depend on some of these ALS-related genes. Thus, by indirectly excluding the patients with these ALS-related genes, previous trials could potentially have excluded responding patients in trials and missed subtle treatment clues.”

The findings raise questions regarding not only the value of currently applied eligibility criteria but also the generalizability of clinical trial results. Investigators think that prediction models could individualize selection and optimize the balance between endpoint heterogeneity and the generalizability of trial outcomes.

When asked how the eligibility criteria should be revised to include more patients, van Eijk explained that the most important consideration is to no longer look at group-level patient characteristics (e.g. all patients with symptoms of more than 24 months, or all patients with a lung function less than 60%), but to look at the characteristics of individual patients (e.g. the combination between duration of symptoms and lung function). The researchers utilized a prediction model, that assesses, for example, the risk for a patient not to complete the trial or not show any change during the follow-up period. The prediction is specific for an individual and is based on the patient’s age, symptom duration, lung function, and functional scores. With the prediction model, investigators can “personally” select patients for trials which would not exclude whole groups of patients, but rather a few individuals allowing more patient participation. This could reduce patient heterogeneity more effectively while balancing generalizability and eligibility.

“In a comparison between patients that are eligible to participate and patients that actually participated, we found some interesting differences,” van Eijk concluded. “The patients that actually choose to participate in trials are more often younger males in a very early stage of the disease. This is a selection process that is not caused by inclusion criteria and is a deeper, more psychological process. This is thus something that is unlikely to change by changing the inclusion criteria. Our future research will aim to identify these psychological factors and to make it easier for patients to participate in trials. For example, we are currently evaluating methods that evaluate drug effects at home. In this manner, patients don't need to visit the clinic anymore and might lower the burden to participate in trials. We hope that more older patients and patients that are in a more severe stage of the disease will participate. By identifying the underlying rationale of a patient to participate (or not to participate) may help us to improve our information and recruitment methods for future trials.”

REFERENCE

van Eijk, R, Westeneng HJ, Nikolakopoulos S, et al. Refining eligibility criteria for amyotrophic lateral sclerosis clinical trials. Neurology. 2019;92(5).

doi

: 10.1212/WNL.0000000000006855.