Date Published: October 4, 2017
Publisher: BioMed Central
Author(s): Michelle M. Nuño, Daniel L. Gillen, Kulwant K. Dosanjh, Jenny Brook, David Elashoff, John M. Ringman, Joshua D. Grill.
Research has revealed that manifest Alzheimer’s disease (AD) dementia is preceded by preclinical and prodromal phases during which pathology is accumulating but function remains intact. This understanding and concern that disease-modifying interventions initiated at the dementia stage may come too late in the neurodegenerative process to be successful has led to a paradigm shift in AD clinical trials. AD trials now enroll patients with mild cognitive impairment (MCI) and persons with no cognitive symptoms. Trial designs are similar to those enrolling dementia participants. We set out to test the hypothesis that attitudes towards trial design features differ among different potential AD trial populations.
We sent a survey composed of 37 items assessing specific trial elements to 246 cognitively normal, MCI, and AD dementia participants at the University of California Los Angeles (UCLA) Alzheimer’s Disease Research Center (ADRC), from whom we received 91 responses (37 cognitively normal, 32 MCI, and 22 dementia). To quantify willingness to enroll, we created three composite scenarios by summing responses and fitting proportional odds models with a binary outcome variable for whether patients were highly willing to participate in low-, moderate-, or high-risk and burden trials.
MCI participants less frequently correctly self-identified their diagnoses than those with dementia or normal cognition. Compared to dementia patients, the odds of participating in a low-risk, low-burden trial were 12% lower for MCI patients (odds ratio (OR) = 0.88, 95% confidence interval (CI) 0.23–3.29) and 70% lower (OR = 0.30, 95% CI 0.08–1.09) for cognitively normal participants. With increasing risk and burden, willingness to enroll decreased and the gap in relative willingness between diagnostic groups increased. In the medium-risk, medium-burden scenario, the estimated OR was 0.64 (95% CI 0.17–2.40) for MCI and 0.21 for the cognitively normal (95% CI 0.06–0.77). In the high-risk, high-burden scenario, the estimated OR indicated reduced willingness for MCI (OR = 0.27, 95% CI 0.06–1.15) and cognitively normal respondents (OR = 0.12, 95% CI 0.03–0.54).
These results suggest that AD trials enrolling predementia populations, especially those requiring frequent visits and implementing biomarker testing procedures, may encounter challenges to enrollment.
The online version of this article (doi:10.1186/s13195-017-0311-5) contains supplementary material, which is available to authorized users.
Alzheimer’s disease (AD) is a progressive neurodegenerative disorder and the most common cause of dementia. A few symptomatic therapies are approved for AD but, as yet, no therapy has been successfully demonstrated to slow the cognitive and functional declines that characterize the disease . Myriad challenges to developing disease-modifying therapies for AD exist. Among these, numerous studies indicate that by the time dementia is diagnosed neurobiological changes have been occurring for a decade or longer . Based on this premise, research diagnostic criteria for mild cognitive impairment (MCI) due to AD  and prodromal AD [4, 5], as well as preclinical  or asymptomatic AD [4, 5], have been proposed for use in earlier disease clinical trials of potential disease-modifying therapies.
Ninety-one participants completed the survey: 22 with a clinical diagnosis of dementia, 32 with MCI, and 37 with normal cognition. One respondent’s consensus diagnosis was cognitive impairment not meeting criteria for dementia or MCI. In analyses assessing attitudes towards trial participation, this patient was included as part of the MCI group. The participant was excluded from analyses of self-reported and consensus diagnosis agreement. Table 1 describes the demographic characteristics for the study respondents stratified by clinical diagnosis. The MCI group had an equal number of males and females, while the cognitively normal and dementia groups had more males. Most participants were white (91.1%). The mean age of the respondents was 73.5 years for those with dementia, 69.9 years for MCI, and 72.5 years for the cognitively normal. Dementia participants tended to be more educated. The majority of participants were married and most lived with a spouse, though the proportion living with a spouse was highest among the dementia group. As expected, most cognitively normal and MCI participants functioned independently, while the majority of dementia participants required assistance with basic or instrumental activities of daily living. Similarly, most cognitively normal and MCI participants reported completing the survey on their own (80.6% and 80.7%, respectively), while a majority of dementia participants reported receiving assistance from their study partner on more than half the questions or completing the survey together with their study partner (90.9%).Table 1Characteristics of survey respondentsParticipant characteristicsConsensus diagnosisCN (n = 37)MCI (n = 32)Dementia (n = 22)Age, mean (SD)72.46 (10.4)69.94 (9.5)73.55 (11.4)Education, mean (SD)17.22 (1.9)16.78 (2.7)19.73 (17.9)Female sex, n (%)16 (43.2%)16 (50.0%)6 (27.3%)Race, n (%) White35 (97.2%)29 (90.6%)18 (81.8%) African-American0 (0.0%)3 (9.4%)1 (4.6%) Asian1 (2.8%)0 (0.0%)2 (9.1%) Other0 (0.0%)0 (0.0%)1 (4.6%)Hispanic, n (%)0 (0.0%)1 (3.1%)1 (4.6%)Marriage status, n (%) Married25 (67.6%)22 (68.8%)17 (77.3%) Widowed6 (16.2%)1 (3.1%)3 (13.6%) Divorced3 (8.1%)6 (18.8%)2 (9.1%) Never married2 (5.4%)3 (9.4%)0 (0.0%) Living as married1 (2.7%)0 (0.0%)0 (0.0%)Partner involvement, n (%) No help29 (80.6%)25 (80.7%)2 (9.1%) Less than half5 (13.9%)1 (3.2%)0 (0%) More than half2 (5.6%)5 (16.1%)13 (59.1%) Together0 (0.0%)0 (0.0%)7 (31.8%)Living situation, n (%) Lives alone7 (18.9%)8 (25.0%)3 (13.6%) Lives with spouse or partner22 (59.5%)20 (62.5%)16 (72.7%) Lives with relative or friend2 (5.4%)0 (0.0%)1 (4.6%) Lives with group4 (10.8%)1 (3.1%)1 (4.6%) Other2 (5.4%)3 (9.4%)1 (4.6%)Independence, n (%) Able to live independently36 (97.3%)31 (96.9%)8 (36.4%) Requires some assistance with complex activities0 (0.0%)1 (3.1%)12 (54.6%) Requires some assistance with basic activities1 (2.7%)0 (0%)1 (4.5%) Completely dependent0 (0.0%)0 (0%)1 (4.5%)“Partner Involvement” refers to how much the study partner assisted the patient in completing the survey“Independence” refers to how well participants can perform daily activitiesCN cognitively normal, MCI mild cognitive impairment, SD standard deviation
The results of this study indicate that the simple adoption of AD dementia trial design features in predementia trials may result in challenging trial recruitment. These recruitment challenges may be most evident in trials that involve more invasive assessment methods (such as PET or lumbar puncture) or more invasive treatments (such as infusions or vaccines). For example, the odds of high willingness to enroll in a low-risk, low-burden trial for MCI participants were estimated to be 12% less (0.88, 95% CI 0.23–3.29) than that of dementia participants. For a high-risk, high-burden trial, however, the odds of high willingness to participate were estimated to be 73% less (0.27, 95% CI 0.06–1.15). The odds of high willingness to participate for cognitively normal participants were estimated to be 70% (0.30, 95% CI 0.08–1.09), 79% (0.21, 95% CI 0.06–0.77) and 88% (0.12, 95% CI 0.03–0.54) less than those of the dementia group in low-, medium-, and high-risk and -burden composite scenarios, respectively.
Recruitment for Alzheimer’s disease clinical trials is often a difficult task. If trials are burdensome, participants and their study partners are less likely to enroll. We found that in exemplary trial scenarios, the willingness to participate was greater for dementia patients compared to MCI and cognitively normal participants. These results suggest that AD dementia clinical trial designs may encounter even more difficult recruitment when used for prodromal and preclinical AD studies. Moreover, these results indicate the need for planning studies to ensure recruitment feasibility for each diagnostic category. Major international efforts to create cohorts that are “trial ready” may address some of these challenges [33–35], but a more thorough understanding of which design elements encourage participation in each potential study population will allow researchers to carefully design trials in ways that optimize recruitment and ensure trial success.