Optimization of Treatment Algorithms for Clinical Depression: Glutamate Antagonists and Transcranial Magnetic Stimulation as Case Examples

Major depressive disorder (MDD) is a very prevalent disorder (1) that is associated with a high degree of burden and substantial disability worldwide (2). Despite many advances in the treatment of depression, treatment outcomes often remain suboptimal, with only one-third of patients achieving symptomatic remission after initial treatment (3). Sustained symptomatic remission has for many years now been the goal of antidepressant treatment (4, 5), although increasing recognition is being given to the need to evaluate functional outcomes as the ultimate treatment goal to improve long-term outcomes (6, 7). There are still many improvements needed in our approach to treating depression to better pair effective treatments with individual patients with the goal of obtaining better treatment outcomes more rapidly; this is the impetus behind the current focus on personalization of treatment in the field.

Algorithms have been helpful in improving treatment outcomes and increasing treatment concordance with evidence-based recommendations and guidelines in both specialty (3, 8–11) and primary care settings (3, 9–13). Algorithms typically recommend a sequenced approach to treatment strategies with an emphasis on providing a treatment trial of adequate dose and duration to sufficiently assess a given treatment’s efficacy (4). Algorithms guide clinicians with respect to a number of parameters regarding antidepressant treatment, such as length of treatment, dose level of medication, frequency of administration and monitoring, and assessment of side effect burden (see Trivedi et al. [8] for an example of a treatment algorithm). However, in practice, utilization of treatment guidelines and algorithms has been fairly slow, contributing to suboptimal utilization of medications (13, 14), use of inadequate dosages and duration of treatment, and use of inadequate strategies for medication titration, switch, and augmentation (15, 16). Thus, an important issue in the field has been development of strategies to enhance provider acceptance and utilization of treatment algorithms.

Measurement-based care (MBC) is a technique that was developed in the Texas Medication Algorithm Project (TMAP) and the Sequenced Treatment Alternatives to Relieve Depression (STAR*D) studies to aid treating physicians in effective implementation of clinical practice guidelines and treatment algorithms. There are four basic steps involved in MBC:

Methods to more accurately select initial or ongoing treatments based on individual biomarkers are currently underway. Until those are realized, the current approach is based upon the clinical integration of information from the patient’s diagnostic and treatment history, current clinical presentation, and anticipated effectiveness, tolerability, safety, and affordability of potential treatment options (17). Once a treatment is selected and administered, adequate monitoring of treatment response is critical. This most effectively occurs with assessment measures that are specific, targeted (i.e., assessing one issue, such as efficacy), psychometrically valid, brief, and easy to administer (18). A wide variety of assessment tools are available to evaluate patient-reported outcomes with respect to depressive symptomatology, side effect burden, tolerability, treatment adherence, and functioning (see Morris et al. [17] for a review). MBC can provide standardized measurement in clinical practice that can improve effective management of patient outcomes. The American Psychiatric Association (APA) guidelines now call for the integration of MBC into routine practice, although this has also been very slow to be realized (18).

Earlier iterations of treatment algorithms focused mostly on effective management of chronic and/or treatment-resistant depression (4), but it is becoming increasingly recognized that the integration of biomarker data and personalized treatment approaches will be essential to future algorithms to maximize selection of the optimal first-line treatment for a given patient. In fact, lessons learned from the conduct of trials evaluating treatment algorithms, such as the STAR*D trial, have reinforced the need for MBC and a personalized approach to treatment. In STAR*D, comparison of treatments following treatment failure with citalopram monotherapy did not identify a clearly superior second-step treatment recommendation; augmentation or switching with a variety of compounds yielded comparable treatment outcomes (19). Similarly, in the Combining Medications to Enhance Depression Outcomes (CO-MED) trial, instigation of treatment with two different combined treatment regimens (bupropion+escitalopram or venlafaxine+mirtazapine) resulted in similar treatment outcomes to escitalopram monotherapy (20). These trials truly underscore the point that a variety of potential beneficial treatments are available, but individual factors may be of greatest importance in the selection and maintenance of any given individual’s treatment. MBC provides the tools needed to adequately monitor a given patient’s unique response to treatment and more importantly provides a road map for next step treatments for patients not achieving the desired treatment goals with previous treatments.

Although the use of MBC is an important strategy to assist in increased utilization of algorithms, another important consideration with respect to treatment guidelines and algorithms is the current focus on the development of novel treatment approaches for depression. The field has clearly seen that no single treatment is universally effective for all patients, and the push for new, effective treatment approaches must continue. When evidence becomes available with respect to promising new treatments, those responsible for developing guidelines and algorithms must consider how and when to incorporate these treatments into treatment recommendations. This requires balancing premature adoption of insufficiently tested treatments with rapid dissemination of those with sufficient evidence supporting their efficacy, safety, and tolerability. The APA practice guidelines for major depressive disorder illustrate how experts in the field navigate this balance (21). Inclusion of novel treatments such as repetitive transcranial magnetic stimulation (rTMS) and some consideration of the benefits of alternative treatment strategies such as exercise demonstrate the integration of novel treatment approaches and consideration of their use by treating physicians. Another example is the burgeoning evidence associated with glutamate antagonists, such as ketamine and metabotropic glutamate receptors (mGluRs), particularly mGluR5. Several lines of converging evidence suggest that these agents produce rapid antidepressant effects. At what point should these agents become part of the clinical treatment repertoire via incorporation into treatment guidelines and algorithms? We review below evidence associated with the efficacy of glutamate antagonists and rTMS as examples of how such considerations are made.

The above descriptions briefly describe data regarding two types of potential antidepressant treatments—rTMS and glutamate antagonists. The former was approved by the FDA as a treatment for depression after two or more decades of research supporting its efficacy and despite at least some concerns regarding how robust of an effect rTMS can produce. We also have some very promising data with clinical trials of glutamate antagonists such as ketamine, balanced with concerns of abuse potential and untoward side effects. The question then becomes, at what point, if at all, do we incorporate this medication into treatment algorithms for use in clinical care? Rush (35) and Schatzberg (36) among others caution the immediate use of ketamine in clinical care for several reasons. Even though the clinical trial data are indeed promising and warrant further evaluation, several adverse effects (e.g., elevated blood pressure, anxiety, dissociative symptoms) should be further evaluated. Furthermore, participants in clinical trials are not representative of the general population and, therefore, clinical trial results are not sufficiently generalizable to move clinical practice without additional investigation in a more representative patient group (i.e., with comorbid medical conditions, taking concomitant medications, etc). In addition, ketamine has produced robust immediate results, but few studies have been conducted to address questions regarding how long it should be administered and what the ideal maintenance treatment might be to sustain remission. Furthermore, dose-finding studies have not yet been conducted and long-term sequelae associated with repeated administration have not yet been determined (28). Thus, while promising, it appears that a number of critical questions remain unanswered and, therefore, guideline developers have not yet encouraged the adoption of this treatment into routine care. Several unanswered questions about the applicability of ketamine in routine practice include: 1) how best to sustain the immediate beneficial effect; 2) concerns about adverse events especially following repeated administrations; 3) evaluation of outcomes in “real-life” participants; 4) most appropriate dosing and intervals for repeated dosing; and 5) longer term consequences like abuse potential and cognitive effects. If these questions can be satisfactorily answered in the next several months or years, perhaps ketamine will become a recommended treatment approach. It is likely that those working on guideline and algorithm development will be closely monitoring ketamine, mGluR5, and other glutamate antagonists and will be eagerly evaluating current research aimed at addressing these lingering questions.

Although it is exciting and hopeful that potential novel treatment approaches are emerging, it is important to balance the need for sufficient evidence of treatment efficacy, in addition to safety and tolerability, with the need to rapidly disseminate the potential utilization of treatments deemed ready for clinical use. Regardless of the treatments added to the clinical armamentarium, it is essential that the field maintain a focus on personalization of treatment to ensure that appropriate treatments are provided to a given patient. This can be accomplished with the use of MBC, which provides both patient and the clinician with patient-specific information regarding their symptoms, side effects, safety, and adherence. In addition, information gleaned from forthcoming trials based on the Research Domain Criteria (RDoC) initiative (40) will be important to consider that may aid in guiding treatments with respect to specific symptoms or domains (e.g., positive and negative valence, cognition). Biomarker driven treatment selection and treatment/outcome monitoring will be essential to future iterations of treatment guidelines that will also assist in personalization of treatment, with the hope of better matching patients with the most beneficial treatment for them.

Tracy L. Greer, Ph.D., Department of Psychiatry, University of Texas Southwestern Medical Center, Dallas, TX

Dr. Greer has received honoraria from and served as a consultant and on a task force for Lundbeck and received honoraria from and served on an advisory board for Takeda.

Madhukar H. Trivedi, M.D., Department of Psychiatry, University of Texas Southwestern Medical Center, Dallas, TX

In the last 12 months, Dr. Trivedi is or has been an advisor/consultant to Alkermes, AstraZeneca, Bristol-Myers Squibb, Cerecor, Concert Pharmaceuticals, Eli Lilly and Company, Forest Pharmaceuticals, Janssen /Johnson and Johnson, Lundbeck, MedAvante, Merck, Mitsubishi Tanabe Pharma Development America, Naurex, Neuronetics, Otsuka Pharmaceuticals, Pamlab, Phoeniz Marketing Solutions, Ridge Diagnostics, Roche, Shire Development, Sunovion, Takeda, and Vivus. In addition, he has received research support from Corcept Therapeutics, the National Institute of Mental Health, and the National Institute on Drug Abuse.