Diversity Inclusion in Clinical Trials Investigating Esketamine for Depression: A Systematic Review

Diversity and Inclusion in Esketamine Clinical Trials

Introduction

• A systematic review was conducted to assess diversity and inclusion in clinical trials for esketamine, a drug approved for treatment-resistant depression (TRD).
• The review highlights the underrepresentation of non-White and Hispanic/Latinx participants in these trials.
• Many studies also fail to address ethnic differences, raising concerns about the generalizability of clinical findings to diverse populations.

Public Health Significance

• The underrepresentation of diverse backgrounds in esketamine clinical trials suggests that clinical findings may not be generalizable to these populations.

Keywords

• esketamine, depression, diversity, ketamine, inclusion

Major Depressive Disorder (MDD)

• MDD affects roughly 8% of adults in the U.S., nearly 21 million people in 2020.
• Individuals with MDD experience significant emotional, somatic, and functional impairments.
• The estimated economic burden of MDD in the U.S. is over $300 billion.

Esketamine

• Esketamine is a new drug approved by the FDA for treatment-resistant depression (TRD).
• Treatment-resistance is defined as a lack of clinical benefit after treatment with at least two different antidepressants at adequate dosage and duration.
• Esketamine is the s-enantiomer of ketamine, a rapid-acting general anesthetic approved in 1970.

Mechanism of Action

• Esketamine's precise mechanism of action as an antidepressive agent is not entirely clear.
• It is proposed to be a nonselective, noncompetitive antagonist that blocks the N-methyl-D-aspartate (NMDA) receptor on γ-aminobutyric acid (GABA) interneurons.
• It also activates the α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptor.
• Activation of the AMPA receptor is thought to enhance brain-derived neurotrophic factor (BDNF) production, which has been found to be diminished in the prefrontal cortex and hippocampus of some individuals with MDD.

Promising Results

• Preliminary results show significant improvement in depressive symptoms and remission rates.
• Improvement in psychosocial functioning.
• Improvement in suicidality.

Demographic Representation in Clinical Trials

• MDD is widespread among varying demographic subgroups, with a higher prevalence in young adults (aged 18–25 years), women, and individuals who report two or more races.
• Multiracial individuals represent less than 5% of participants in MDD treatment trials.
• Black individuals were represented at one fifth the rate of White or European Americans, and Hispanic/Latinx individuals were represented at one seventh the rate of White individuals.
• Only 55% of RCTs reported participant race/ethnicity information from 2005 to 2016.
• Racial disparities are present in nearly all facets of MDD, including risk factors, symptom presentation and severity, and treatment availability.
• Gender discrepancies exist in MDD symptom presentation.
• Demographic factors may influence antidepressant response.
• Genetic ancestry may partially predict treatment response to antidepressants.
• Socioeconomic factors, clinical factors, and race have been shown to significantly predict treatment response.
• Sociodemographic factors such as age, sex, and genetic ancestry represent biological differences.
• Other sociodemographic factors, such as race and ethnicity, represent social constructs.
• These factors tend to serve as proxies for larger genetic, environmental, social, and systemic factors.

Study Objective

• Examine to what extent diverse groups have been represented in RCTs for esketamine for the treatment of depression.

Materials and Methods

• Systematic review was executed according to Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines.

Search Strategy

• Pubmed and Embase were searched for peer-reviewed articles in the English language on April 8, 2022.
• The search process was conducted by one reviewer with guidance by a coauthor.
• The quality assessment was conducted by two reviewers.
• Search terms used: “depression” OR “major depressive” AND “randomized controlled” OR “randomized trial” OR “clinical trial” AND “esketamine.”

Eligibility Criteria

• Involved a unique sample of human participants.
• The intent of the research was to investigate the efficacy of esketamine specifically for the treatment of MDD and/or TRD.
• Involved a randomized control trial (RCT) design.
• Written in the English language.

Exclusion Criteria

• Did not involve an independent RCT (i.e., no meta-analyses, reviews, or commentaries).
• Investigated esketamine for disorders other than MDD.
• Completed with animal subjects.

Data Extraction and Synthesis

• The following information was extracted from each included article:
  • Funding source
  • Sample size, effect size, and completion of power analysis
  • Reporting and completeness of sample demographics including age in years, sex/gender, race/ethnicity, and socioeconomic status (SES)
  • Sample demographic information reported
  • Presence of subgroup analyses by each demographic group
  • Acknowledgment of demographic inclusion in the limitations section of the article’s discussion
• Sex (a biological variable) and gender (a social construct) represent distinct constructs.
• Studies that reported inclusion of both males and females and, if applicable, nonbinary or other gender identities, were categorized as full reporting.
• Studies that only reported race, only reported ethnicity, or only reported the race of a subset of the sample were categorized as partial reporting.

Quality Assessment

• Two reviewers completed a quality assessment of the final included studies.
• Quality assessment was guided by Downs and Black’s (1998) 27-item quality review checklist.
• This checklist assesses the quality of each study’s reporting methods, external and internal validity, bias, sample independence, and power.
• Each reviewer assessed each study independently, and interrater agreement was calculated using Cohen’s κ (M = .94).
• Studies were given a quality index score (QIS).

Transparency and Openness

• This article describes in detail the data exclusions, the search strategy, and exact results.
• All included citations in the review are available through Pubmed.
• All data calculations were completed in Microsoft Excel.
• This study’s design and analyses were not preregistered.

Results

Search Results

• The search strategy yielded 408 articles.
• After removing files that were not published studies, 220 studies were screened.
• The initial abstract review excluded 188 studies due to not meeting inclusion criteria.
• The remaining 32 studies were reviewed in full, and an additional 21 studies were excluded primarily due to secondary analysis on duplicate samples.
• The remaining 11 studies were included.

Quality Scores

• The QIS ranged from .81 to .89 out of a possible 1 (M = .85).
• Study reporting (M = .91), external validity (M = .33), bias (M = .90), and confounding (M = .91) were all calculated.
• All studies reported completion of a priori power analyses with power ranging from 80% to 95%.
• Interrater agreement was strong as measured by Cohen’s κ (M = .94).

Summary of Included Articles

• A total of 11 articles met inclusion criteria.
• These studies cumulatively included 1,878 adult participants recruited from sites in 27 countries.
• The participants ranged in age from 18 to 75+ years, with an average age of 46 years.
• Most studies (90.9%) received funding support from Janssen Research and Development, LLC.
• All (100%) included studies reported on the inclusion of female participants.
• Females were well represented in the included clinical trials, accounting for 62% of all participants.
• Nine studies included both male and female sex/gender reporting; however, two studies only addressed female participation.
• The male population represented about 38% of the total participant population.
• No studies reported the inclusion of transgender or nonbinary individuals.
• Further, all (100%) studies executed power analyses to determine adequate sample size.
• The average power was 88.9%, and studies ranged in power from 80% to 95%.
• Additionally, 10 studies (90.9%) reported some measure of effect size.

Racial and Ethnic Information

• None of the studies reported full racial and ethnic information, the majority reported only partial information (90.9%), and one did not report any racial or ethnic information (9.1%).
• Studies that reported partial information all included racial reporting but failed to include ethnic information about their participants.
• Of the 10 studies that reported race, White individuals represented the majority of participants (71.1%).
• The Asian population was the second highest represented group, accounting for roughly 14% of all participants.
• Black participants (6.5%).
• Multiracial/multiethnic participants (1.9%).
• Two American Indian/Alaskan Native participants and no native Hawaiian or other Pacific Islanders were included.
• No study reported ethnic information, therefore Hispanic/Latinx inclusion cannot be ascertained.

Socioeconomic Status (SES)

• Two (18.2%) studies reported on SES.
• Both articles that addressed SES recorded “any type of employment,” defined as: “any category containing ‘employed,’ sheltered work, housewife or dependent husband, and student”.
• “any type of unemployment,” defined as “including any category containing ‘unemployed’”.
• “other,” defined as retired or no information provided.
• Between these two studies, 57.5% reported employment, 33.1% reported unemployment, and 9.4% reported “other.”

Subgroup Analysis

• Nearly half (45.4%) of the studies included some form of subgroup analysis.
• One study completed subgroup analysis on different age groups only; two studies completed subgroup analyses on age, race, sex, and region; one study completed subgroup analyses on age, sex, and region; and one on sex, age, and race.

Limitations

• Two (18.2%) studies addressed demographic factors in their limitation sections.
• One addressed inadequate male representation and both addressed low representation of non-White participants.

Discussion

• Across studies, females were well represented, non-White and Hispanic/Latinx participants were underrepresented, and too few studies reported on socioeconomic factors to draw meaningful conclusions about representation of various SES groupings.
• The minority of studies completed subgroup analyses on different sociodemographic factors, and even fewer addressed lack of diversity in their study’s limitations section.
• This body of literature exhibited a strength in its reporting and inclusion of female participants, as females comprised 62% of the collapsed samples.
• This contrasts with prior pharmacological research involving depression that did not include adequate female reporting or representation (Weinberger et al., 2010).
• Females were included at an overall higher rate than males, which is in line with the sex prevalence discrepancy in MDD (National Institute for Mental Health, 2022).
• An important limitation involves underutilization of subgroup analyses across sexes.
• Research suggests that males and females may respond differently to antidepressant treatments due to sex differences in the pharmacokinetics and pharmacodynamics of these drugs.
• Subgroup analyses are an important method for identifying and understanding these differences not only across sex, but across other demographic groups, as well.
• Many clinical trials lack sufficient power to complete these analyses.
• No articles reported the inclusion of transgender or nonbinary individuals, even though transgender individuals are nearly four times more likely than cisgender individuals to experience depression.
• Some included studies used terminology such as “women” or “men,” whereas other used “female” or “male.”
• None of the included studies addressed biological sex and gender identity as separate demographic characteristics, nor did they specify which construct they were referring to in their reporting.
• Much like prior MDD trials, most studies included in the present review involved middle-aged individuals (mean age = 46 years).
• One study was specifically targeted at older adults (Ochs-Ross et al., 2020), which is encouraging, considering that depression tends to increase with older age (Luppa et al., 2012).
• White individuals represented the significant majority of reported participants in these trials.
• Black individuals were generally underrepresented in the given population (6.5%), as were Hispanic/Latinx (0%), multiracial (<2%), American Indian/Alaskan Natives (<1%), and Native Hawaiian/Pacific Islanders (0%).
• Hispanic/Latinx individuals in particular may have been involved in a greater capacity than is noted, as no study reported on Hispanic ethnicity.
• A relevant strength of this body of research is the inclusion of geographically diverse recruitment sites.
• Nearly 30 countries were included.
• Only a small portion (18.2%) of the present studies included such recognition, highlighting an area for future improvement.
• Sample sizes needed to assess effects across two subgroups are often much larger than those needed to understand the overall treatment effect.
• An RCT with a power of 80% for an overall treatment effect had less than 30% power to detect a subgroup effect of the same magnitude.
• Sample sizes would need to increase nearly fourfold for subgroup analyses to reach similar power as the overall treatment effect.
• Bierer et al. (2021) suggest that pooled analyses across a series of trials for product development may be one such method for reaching adequate power to conduct reliable subgroup analyses.
• Subgroup analyses are dependent, however, on initial collection of relevant data, further echoing the need to consider and collect demographic and socioeconomic factors in this body of research.

Limitations of the Review

• Only studies completed in the English language were included for review.
• The search strategy was limited to two databases.
• The search process was conducted by one reviewer.

Conclusions

• This systematic review investigated diversity inclusion in 11 RCTs studying the use of esketamine for MDD.
• Articles typically reported some degree of sex/gender and racial reporting, but often did not address ethnicity.
• Female participants were well represented, although non-White participants were largely underrepresented.
• Most studies did not measure SES, and many did not execute subgroup analyses, potentially due to inadequate sample size.
• Future research should consider increasing sample sizes to accommodate these analyses.
• Increased efforts should be directed toward ensuring appropriate and representative recruitment in clinical trials for esketamine.
• This information would assist minority populations and their prescribing physicians in making informed medical decisions about the safety and/or potential risks of using esketamine based on the current literature base.