Download the full paper (PDF, 38 pages)
Abstract
The tricyclic antidepressant (TCA) group of drugs have been in use for 70 years and are effective for severe and melancholic depression, often outperforming newer agents. They are also widely used for other disorders such as headache and pain. This narrative review updates the pharmacology, interactions, adverse effects, therapeutic drug monitoring, safety, and principal clinical indications of commonly used TCAs. This heterogenous group highlights the importance of the principles on which the concept of Neuroscience Based Nomenclature (NbN) is founded, because that informs mechanism-based prescribing.
They are a pharmacologically heterogeneous group and I strongly advocate for speeding the adoption of NbN to emphasise their pharmacodynamics, not their shared fused-three-ring structure. Their optimal utilization has been hampered by insufficient knowledge and teaching, and enduring misconceptions about their pharmacological properties. The promoted safety advantages of newer drugs are overemphasised, mostly based on low-quality data. The supposed safety disadvantages of TCAs are of lesser magnitude and consequence than is widely taught and accepted.
I also draw particular attention to the serious inadequacies of RCTs, especially problems related to generalising their results to specialist clinical populations (external validity) — a problem that is amplified by the meta-analyses which dominate guidelines to a disproportionate extent. Severely ill patients (e.g., with melancholia, suicidality, psychosis), are excluded from these studies, which have thus been distorted by a greater proportion of patients with non-melancholic illness. I explicate NbN principles, and give greater consideration (relative to RCTs), to decades of worldwide clinical experience. The epistemological value and validity of clinical experience is supported by Pearl’s ‘Causal Inference’ theory. Specific members of this group should be considered more often, and earlier, in treatment algorithms, especially for melancholic and severe illnesses, and those who have experienced inadequate improvement with two different types of antidepressants.
Contents
- Abstract — 1
- Contents — 2
- 1 Introduction and overview of current relevance — 4
- 1.2 The decline of TCAs, poor research, blinkered theory — 5
- Key points 1: RCT, and non-RCT, evidence — 6
- 1.3 Safety issues — 7
- 1.3.2 Therapeutic drug monitoring: neglected — 7
- 2 Neuroscience based nomenclature (NbN): the importance of receptor profiles — 8
- 2.1 Beyond receptor profiles — 9
- Key points 2 Combined treatments — 9
- 2.2 Interpreting receptor affinities — 10
- Table 1. Relationships between hSERT and various serotonin mediated effects (SMEs) — 10
- 3 Individual drugs — 11
- 3.1 Less commonly, prescribed members of this group — 11
- 3.2 Amitriptyline (AMI) — 11
- 3.3 Nortriptyline (NTP) — 12
- Key points 3: Properties of an ideal drug (e.g. NTP) — 12
- 3.4 Imipramine (IMI) — 12
- 3.5 Desipramine (DMI) — 12
- 3.6 Doxepin (DXP) and Trimipramine (TMI) — 13
- 3.7 Clomipramine — 13
- Key points 4 TCAs, serotonin receptors and adverse effects — 13
- 4 Mechanism of Action — 14
- 4.1 In general — 14
- 4.1.1 Inhibition of neurotransmitter reuptake — 14
- 4.1.2 Receptor antagonism profiles — 14
- 4.1.3 Multi-receptor actions: ‘Dirty’ drugs — 15
- Key points 5: Dirty drugs — 15
- 4.1.4 Mechanisms of SRI and NRI effects — 15
- 4.1.5 Sleep–wake regulation — 16
- 4.1.6 Cardiac electrophysiology and ion channels — 16
- 4.1.7 Emotional blunting and cognition — 17
- 4.1.8 Long-term sexual side-effects — 17
- 5 Side Effects — 17
- 5.1 Antimuscarinic effects — 17
- Key points 6 Antimuscarinic effects — 17
- 5.2 Sedation and weight gain — 18
- 5.3 Sexual side-effects — 18
- 5.4 Orthostatic hypotension — 18
- 6 Safety profile — 18
- 6.1 Drug combinations and switching risks — 18
- 6.2.1 Cardiovascular risks: QT prolongation — 19
- 6.2.2 Practical mitigation in psychiatric practice — 19
- 6.3 Toxicity in overdose — 19
- Key points 7 Suicide risk and toxicity — 20
- 7. Clinical Indications — 20
- 7.1.1 Unipolar major depressive disorder — 20
- 7.1.2 Bipolar I and II major depressive disorder — 21
- 7.2 Anxiety disorders — 22
- 7.2.1 Panic disorder — 22
- 7.2.2 Generalized anxiety disorder — 22
- 7.2.3 Obsessive-compulsive disorder — 22
- 7.3 Insomnia — 22
- 7.4 Other uses — 22
- Table 3 from [7] — 24
- 9 When might TCAs be a better choice — 24
- 10 ‘Desert Island disc’ selection — 24
- Perspective and concluding comments — 25
- Acknowledgements — 25
- References — 27
Summary
Perspective and concluding comments
This review summarises the increasing evidence of this drug-group’s diverse usefulness and their continuing relevance for the treatment of severe depression, as well as their many other uses and properties. Some of them are valuable combining and augmenting agents. It illustrates that the lessons of history are easily forgotten, as illustrated by the notable failure to cite a key review by Kragh-Sørensen, which consolidates the evidence for the superiority of CMI, or to be aware of key observations made about these drugs in the early 1960s, especially the possible added therapeutic benefit of multiple receptor (dirty) actions. Those early observations might have led to more incisive research, had they been noted and remembered.
It is emphasised that an important perspective in assessing and interpreting the evidence and experience accumulated over 70 years is the emerging view in epistemological theory that RCTs represent a narrow and inadequate approach to understanding causes and mechanisms. That understanding is a sine qua none of heuristically directed scientific research. RCTs have contributed almost nothing to that central cause-mechanism-effect requirement of science, as attested to by the lack of progress in drug investigation research over the last 70 years. The breakthroughs that have occurred have been almost entirely the result of clinical research and experimentation, not from RCTs.
NbN principles and science research data have been slow to permeate clinical practice. Outdated information about TCAs (and MAOIs) persists in psychopharmacology teaching programs and texts — with a few notable exceptions. Misinformation is amplified by the computerised drug-information prescription and warning systems which are now ubiquitous in doctors’ practice software: these contain multiple distortions and errors which may then be interpreted by people who are not appropriately experienced or qualified [232].
It is important to treat the illness to full remission and not overvalue rare negative outcomes that may be only marginally different between treatments. Associative epidemiological studies, which are frequently misleading, have been widely used to infer properties relating to toxicity and adverse effects, but they are a weak methodology and are epistemologically unsound because of an absence of cause-mechanism-effect grounded methodology (Cf. Pearl): they have led to conclusions of poor reliability and clinical usefulness. This has happened in parallel with over-emphasis on adverse effects, insufficiently balanced by recognition of the major morbidity caused by serious depression, which is ameliorated by effective treatment. This reflects generally poor risk assessment, and pervasive risk aversion: the unrealistic goal of finding effective antidepressants that have no side effects is a pipe-dream. As a much respected professor of pharmacology taught me: if someone claims a drug that has no side effects, you can be sure that it will have no benefits either.
It remains important for trainees and early career psychiatrists to update their knowledge of this group whilst recognizing that some current texts, guidelines, and recommendations, are inadequate and outdated.