Clinically Meaningful Gaps Between Clinical Trials and Patient Treatment

In this editorial, we note issues in the psychopharmacological treatment of persons with psychosis, which are usually ignored in regulatory clinical trials and are inadequately addressed in subsequent studies. Virtually all the data to date relate to antipsychotic medications, which have their initiating action at the D2 dopamine receptor.

1. A dose range is known, but often this range defines neither the minimally effective nor maximally tolerated dose. Drug dosing information from regulatory studies typically provides information for a dose range based on limited exploration in the clinical trials of sponsors. These studies do not fully explore optimal dosing and do not provide information at the individual patient level.

2. If a therapeutic response is inadequate, the clinician is not informed as to potential benefit/risk of reducing or increasing dose or switching to another antipsychotic drug. None of these actions is supported by sufficient empirical evidence, and a thought experiment can answer the question as to increased adverse effects associated with higher doses or the use of multiple antipsychotic drugs. Meta-analyses have generally failed to support an advantage in adding a second antipsychotic drug. The CATIE trial,¹ and now presented as a report by Rene Káhn at SIRS 2018, the OPTIMISE first-episode trial, strongly suggest switching has no therapeutic advantage.

3. Patients may be considered poor responders based on Positive and Negative Syndrome Scale total scores or the overall clinical picture. However, we believe that it is crucial to judge response only on the symptom domains for which the antipsychotic drug has purported efficacy. For example, negative symptoms and cognitive impairment and their functional consequences are not addressed by antipsychotic medication and should not be the basis for antipsychotic therapeutic decisions except to avoid drugs with adverse effects on these aspects of illness.

4. Effectiveness is not the same as efficacy. Using time on the drug as a primary outcome measure confounds therapeutic benefit with reduced adverse effects that are dysphoric on a daily basis (in contrast to metabolic issues that may be profound over time, but don’t bother the patient “at the moment”).

5. Since the majority of antipsychotic drugs have the same mechanism of action, without significant between drug differences in efficacy (except for clozapine), the adverse effect profile of a particular agent is the critical issue, which should be taken into account when working with a patient on what medication should be prescribed.

6. Clozapine has modest but superior efficacy in moderate as well as in severe treatment-resistant patients. Most guidelines call for 2 trials with different antipsychotic drugs lasting at least 6 weeks before the initiation of a clozapine treatment trial. The recent TRRIP report² suggested “adequate” duration in their recommendation for when to consider a clozapine trial, but the clinician gets a strong signal from the previous history of response and response in the first 2 weeks of current medication. There seems little reason for a potential clozapine responder to be required to have two sustained periods of psychotic symptoms before qualifying for a clozapine trial. In addition, the clinician can keep in mind the established benefit regarding aggression and suicide, strong reasons to avoid unnecessary delay in clozapine treatment initiation.

7. Not all patients experience the clozapine advantage and the metabolic burden and other difficult to manage side-effects suggest the need to discontinue clozapine if benefits are not observed over a sustained period. The exact duration for an adequate clozapine trial is not clear, but benefits accrue gradually over weeks and time is essential for cross-titration between clozapine and the antipsychotic drug being replaced and determining whether clozapine has efficacy in each individual case.

8. It is important for prescribing doctors to understand clinical trial results and the term “pseudospecificity.” With the introduction of second-generation antipsychotic medications (SGAs), a common view developed that the field finally had antipsychotic drugs with efficacy for impaired cognition and negative symptoms. Study reports that seemed to support this view did not use a clinical trial design for an efficacy indication for these aspects of schizophrenia. Almost any SGA, with their diminished risk for extrapyramidal symptoms, given at appropriate dose, will be better than haloperidol at usual doses, since the latter drug can cause secondary negative symptoms and interfere with cognitive measures. Note that regulatory bodies have not granted a superior efficacy claim to any of the antipsychotic drugs for cognition or persistent or primary negative symptoms. The required design to show efficacy can be found in the following citations.^3–5

… and now for our number 1 pet peeve. Relapse is a trump card dominating every consideration of antipsychotic pharmacotherapy. All patients with schizophrenia are presumed to be in need of continuous therapy for years/maybe life, but the clinician deals with individuals, most of whom will not take drugs continuously forever, most for not even a year. Patients may believe that taking an antipsychotic drug is required for clinical care and may mislead their doctors regarding adherence or may even stop their clinic visits. It would help if we discontinued the use of the term “relapse” and used the term “exacerbation.” The term “exacerbation” is more easily used as a dimension, with severity to be understood, rather than relapse, which is presumed to be severe.

There are antipsychotic drug alternatives for the nonadherent patient. Long acting injectable antipsychotic medications addresses adherence if the patient agrees. But for patients who do not want to take antipsychotic drugs, doctors can work with a targeted drug strategy and attempt to engage the patient in the advantages of recognizing signs of impending exacerbation and reach an agreement of how to intervene at that point. This could include benzodiazepines or similar drugs if early warning signs are stress/anxiety/disturbed sleep based. This has been ignored as a clinical trial option with the exception of one small trial at the Maryland Psychiatric Research Center with fluphenazine and diazepam being of equivalent efficacy and both superior to placebo at this early stage of exacerbation. However, the trump card effect has largely taken targeted drug approaches off the table, despite the fact that the targeted approach would be a desirable option for patients who decline continuous medication. It provides a context for fully integrated clinical care with a focus on the early identification of exacerbations where treatment intervention is most effective. Adverse effect advantages are expected if prolonged periods of off medication are achieved.

The above discussion does not deal with many other tasks for the prescribing doctor, such as the consideration of drug–drug interactions and the profoundly important integration of long-term adverse effects when initiating treatment. Just as tardive dyskinesia should have been influential in reducing dosing with first-generation antipsychotic drugs, the assessment of general health and risk factors for diabetes, cardiovascular disease, etc. need to be considered from the beginning of antipsychotic treatment. This should involve individualized judgments regarding the risk of adverse effects differentially associated with the various antipsychotic drugs, the likely severity of symptom exacerbations and whether they would be detected early enough for intervention if the patient is nonadherent or a targeted strategy is in place.