Two treatments that did not work do not mean you are out of options. They also do not mean the next choice has to be stronger, stranger, or picked at random.
The next useful step may be a small one. It may be to correct the diagnosis, make an earlier treatment a fair one, add therapy, change medicines, or move sooner to a specialty treatment. The right path depends on what has happened so far.
Key takeaways
- “Treatment-resistant depression” is a useful label, but there is no single agreed definition of it.
- Before adding another treatment, your team should check the diagnosis, your safety, how well earlier treatments were really given, sleep, substances, medical causes, and practical barriers.
- Switching, adding a second medicine, TMS, ECT, esketamine, and off-label IV ketamine each solve different problems. They are not one ranked list.
What the label means and what it misses
A common definition of treatment-resistant depression, or TRD, is depression that has not improved enough after at least two fair antidepressant trials. Regulators in the United States and Europe use a version of this two-trial rule. By that measure, at least 30 percent of people with depression would qualify.1
But there is no agreed definition. Studies and clinics do not always mean the same thing by “fair.” They set different rules for dose, for how long a medicine was taken, and for whether the person could take it as prescribed. They also differ on what counts as success. Some count partial improvement. Others count only full remission, which means symptoms are essentially gone.1
The label is a starting point, not a verdict.
Some people meet a research definition after two prescriptions. Others have not really had two fair trials yet. Side effects may have ended a treatment early. The dose or the time may have been too small. A gap in care may have cut a trial short.
Clinicians call this pseudoresistance. The word sounds like it doubts you, but it does not mean the symptoms are fake. It means a fair test of the treatment has not happened yet, because something is in the way. This is common, and it is one of the main reasons people carry a TRD label without meeting the definition.1
Check pseudoresistance without blaming the patient
| Area to check | What the team is trying to learn |
|---|---|
| Diagnosis | Is this unipolar depression, bipolar depression, trauma-related illness, grief, substance-related illness, depression with psychosis, or another condition? |
| Current safety | Is there suicidal thinking, psychosis, catatonia, trouble eating or drinking, or fast-dropping function that changes the order of care? |
| Treatment adequacy | Was the dose high enough, taken long enough, and taken steadily enough to be a fair trial, and was benefit measured in both symptoms and daily function? |
| Tolerability | Did the treatment fail, or did side effects make it impossible to keep taking? |
| Sleep | Could sleep apnea, insomnia, a shifted body clock, restless legs, or another sleep problem be keeping symptoms going? |
| Medical causes | Do thyroid disease, low iron, pain, hormone changes, a nerve or brain condition, or inflammation from a specific illness need their own workup? |
| Medicines and substances | Could alcohol, cannabis, stimulants, sedatives, steroids, or another prescribed medicine be making mood, sleep, or thinking worse? |
| Access and daily life | Did cost, transportation, work, housing, caregiving, language, pharmacy supply, or fear make the plan hard to follow? |
| Psychosocial threat | Is abuse, unsafe housing, discrimination, legal trouble, or another ongoing threat larger than what treatment alone can fix? |
Catatonia, listed above, is a state in which a person may stop moving, speaking, eating, or drinking normally. It needs urgent attention, and it changes the order of care.
“I could not take it every day” is clinical information, not a moral failure. A plan that does not fit your life is not a good plan yet.
No response, partial response, and poor tolerance lead to different talks
No response means little real change in symptoms or in daily function. That may favor a switch, a fresh look at the diagnosis, or a treatment that works a different way.
Partial response means something got better but important symptoms are still there. If you tolerate the medicine well, the team may suggest augmentation. That means keeping the medicine that is helping and adding a second one, so you keep the gain you have and add another route toward remission.
Poor tolerance means side effects are the main reason you could not continue. A second medicine may just add burden. A switch, or an option that is not a medicine, may make more sense.
These are tendencies, not rules. Urgency, past response, pregnancy, other conditions, drug interactions, cost, and your own preference can all change the choice.
The decision map
| Decision point | Possible next move | Reason to pause and reconsider |
|---|---|---|
| Diagnosis or safety is uncertain | Recheck diagnosis, suicide risk, bipolar signs, psychosis, catatonia, substances, sleep, and medical causes | A treatment aimed at unipolar depression may be unsafe, or too slow, for the real problem |
| Prior trial was not adequate | Fix the dose, duration, barrier to taking it, interaction, or measurement, when that is safe | More of a poorly tolerated or wrong treatment is not optimization |
| Psychotherapy was absent or limited | Add or improve an evidence-based therapy matched to the problem | Access, fit, and readiness all matter |
| Little benefit and tolerability is poor | Switch within the class or outside it | There is no rule that a within-class switch must always come first |
| Partial benefit and tolerability is acceptable | Consider augmentation or combination treatment | Added interactions, weight and metabolic effects, movement effects, sedation, or monitoring may outweigh the benefit |
| Severe, urgent, psychotic, catatonic, or repeatedly unresponsive illness | Refer for ECT, TMS, esketamine, off-label IV ketamine, or another specialty evaluation as appropriate | Each option differs in evidence, speed, burden, contraindications, and access |
| Any step | Measure symptoms, function, side effects, and safety; revisit goals and diagnosis | A lower score without a restored life, or without safety, is not enough |
CANMAT and VA/DoD both support either switching or augmentation after a poor response. Neither sets one required order for every adult.2 NICE takes the same view. Its guideline tells clinicians to make a shared decision with the person. It adds that if someone wants to try a combination of medicines, the clinician should consider referral to a specialist mental health service, or advice from a specialist.3
Partial benefit with good tolerance can favor augmentation. No benefit, or poor tolerance, can favor switching. Your choice belongs at every branch.
Current medication status: approval is not the same as fit
The table below reflects U.S. regulatory status as of September 7, 2026. “FDA-approved adjunct” means the medicine is approved to be added to an antidepressant for adults with major depressive disorder. It does not mean it is right for you.
| Strategy | U.S. status, evidence position, and decision factors |
|---|---|
| Aripiprazole | U.S. status: FDA-approved adjunct for adult MDD Evidence and purpose: Strong guideline support for augmentation Burdens and decision factors: Restlessness, movement effects, weight and metabolic change, sedation or activation, interactions |
| Brexpiprazole | U.S. status: FDA-approved adjunct for adult MDD Evidence and purpose: Strong guideline support for augmentation Burdens and decision factors: Weight and metabolic effects, restlessness, sedation, interactions |
| Cariprazine | U.S. status: FDA-approved adjunct for adult MDD Evidence and purpose: Current adjunctive option with randomized-trial support Burdens and decision factors: Restlessness, movement effects, and side effects that can appear late, because one breakdown product stays in the body for weeks |
| Quetiapine XR | U.S. status: FDA-approved adjunct for adult MDD Evidence and purpose: Evidence-supported augmentation Burdens and decision factors: Sedation, blood pressure drops on standing, weight and metabolic effects, heart and interaction review |
| Lumateperone | U.S. status: FDA-approved adjunct for adult MDD since November 2025 Evidence and purpose: Newer approved adjunct; less long-term MDD experience than older agents Burdens and decision factors: Sedation, dizziness, metabolic and interaction review |
| Olanzapine plus fluoxetine | U.S. status: FDA-approved for the acute treatment of treatment-resistant depression, either as the fixed combination or as olanzapine labeled for use together with fluoxetine. Olanzapine on its own is not approved for this use Evidence and purpose: Specific TRD indication for people who did not respond to two fair antidepressant trials in the current episode Burdens and decision factors: Substantial weight and metabolic burden, sedation, interaction review |
| Lithium | U.S. status: Off-label augmentation for unipolar MDD Evidence and purpose: Longstanding guideline-supported option; may matter most when suicide risk is part of specialist planning Burdens and decision factors: Blood level checks, kidney and thyroid monitoring, care with hydration and salt, interactions, pregnancy |
| Liothyronine, or T3 | U.S. status: Off-label augmentation for unipolar MDD Evidence and purpose: Older evidence and guideline support for selected patients Burdens and decision factors: Thyroid and heart assessment, bone and symptom monitoring, interactions |
| Bupropion added to another antidepressant | U.S. status: Off-label combination Evidence and purpose: Common in practice, but evidence is less consistent than for the approved add-on antipsychotics Burdens and decision factors: Blood pressure, seizure risk, sleep or anxiety effects, CYP2D6 interactions |
Labels also carry boxed warnings and drug-specific cautions that cannot fit in a table. A label answers one question: is this use approved. It does not rank the options against each other, and it does not replace a personal review of risk.4
Devices and rapid-acting options in context
ECT is a medical procedure done under general anesthesia. A controlled electrical stimulus causes a brief, therapeutic seizure. It can move earlier when depression is severe, psychotic, catatonic, life-threatening, or needs a fast response. It also carries anesthesia risks, medical risks, and effects on memory and thinking. Modern ECT for Depression: Procedure, Benefits, Memory, and Safety gives the full explanation.
TMS uses magnetic pulses delivered at the scalp. In its usual forms it does not require general anesthesia and does not cause a seizure. Treatment usually means repeated weekday visits, though protocols vary. It is often considered after medicine has not helped enough, especially when avoiding whole-body side effects matters.
The regulatory wording for TMS differs from the wording for medicines. TMS devices are not “FDA-approved.” They are FDA-cleared through the 510(k) route as Class II devices. FDA’s classification covers treating major depressive disorder in people “who have failed to achieve satisfactory improvement from one prior antidepressant medication at or above the minimal effective dose and duration in the current episode,” which means one prior antidepressant, not two.5
Esketamine nasal spray, sold as Spravato, is FDA-approved for adults with TRD, either on its own or together with an oral antidepressant. That “on its own” option is newer, and it applies only to the TRD use. There is a second, separate approval for depressive symptoms in adults with major depressive disorder who have acute suicidal thoughts or behavior, and for that use the label requires an oral antidepressant alongside it.
The label is also clear about what esketamine has not been shown to do. It states that the drug has not been shown to prevent suicide or reduce suicidal thoughts or behavior, and that using it does not remove the need for hospitalization when that is warranted. It must be given through a restricted safety program, with monitoring for sedation, dissociation, slowed breathing, and raised blood pressure, and observation for at least two hours after each dose.6
IV ketamine for depression is off-label. It may act quickly for some people. But the formulation, dose, setting, monitoring, and access all differ from labeled esketamine, as does how long benefit lasts. Neither treatment is a cure, and neither replaces urgent safety care.
| Option | Strongest use case, usual burden, and main limits |
|---|---|
| ECT | Strongest use case: Severe, psychotic, catatonic, urgent, or highly treatment-resistant depression Usual burden: Anesthesia, transport, repeated procedures, cognitive monitoring, maintenance planning Main limits: Memory and medical risks; access and stigma |
| TMS | Strongest use case: Nonpsychotic depression after a poor response, especially when an outpatient option without whole-body side effects is preferred Usual burden: Frequent outpatient visits; no anesthesia in usual forms Main limits: Time, access, and variable response; protocol matters |
| Esketamine | Strongest use case: Labeled adult TRD, plus a separate acute-suicidal-symptom use that requires an oral antidepressant Usual burden: In-clinic dosing and post-dose observation under a restricted program Main limits: Dissociation, sedation, blood pressure, misuse risk, maintenance and cost |
| IV ketamine | Strongest use case: Specialist off-label rapid-acting treatment in selected patients Usual burden: IV access, clinic monitoring, repeated visits Main limits: Off-label status, protocol variation, how long benefit lasts, medical and misuse risk |
Raw response percentages from separate trials should not be used to rank these treatments. The patients, outcomes, timing, and prior failures all differ.
Pharmacogenomic testing stays in a supporting role
Pharmacogenomic testing, sometimes called PGx, looks at gene variants that affect how your body processes certain medicines. It is worth being precise about what it can do.
CPIC, the group that writes prescribing guidance for these genes, gives dosing advice for certain antidepressants when a relevant genotype is already known. It does not tell a clinician which antidepressant will work for you. CPIC also reviewed two genes often sold on commercial panels, SLC6A4 and HTR2A, and concluded that the data “do not support their clinical use in antidepressant prescribing.”7 It makes no prescribing recommendation based on them, because the evidence is mixed or too thin.
The VA/DoD depression guideline reached a matching conclusion in its recommendation 13: there is insufficient evidence to recommend for or against pharmacogenetic testing to help guide the choice of antidepressants.7
So the useful question is narrow. A well-established gene-drug interaction may help explain unusual drug levels, unexpected side effects, or an unusual dose for one medicine. A colorful commercial report is not a diagnosis, and it does not replace your history.
Low-dose naltrexone stays in an experimental lane
Low-dose naltrexone, or LDN, is off-label for depression. It does not have an established place beside ECT, TMS, or the FDA-approved TRD treatments.
The evidence is not just thin. So far it is negative. The one randomized, placebo-controlled trial of LDN added to antidepressant treatment in major depression found no difference from placebo: 37 patients, and after 12 weeks depression scores fell by a similar amount in both groups.8 A trial that small cannot rule out a benefit in some other group of patients. It also cannot be used to claim one.
Low-dose naltrexone in psychiatry explains the hypothesis and the limits of the evidence. It should not be a required step before the treatments above.
What STAR*D really adds
STAR*D followed a large, broadly inclusive group of outpatients through several treatment steps. Some participants did reach remission at later steps. But remission became less common at each level, and more people left the study at each level. No single switch or add-on option clearly won.9
The often-repeated claim that about 67 percent of people eventually remitted needs context. The study’s own report calls that figure a theoretical cumulative estimate. Two things went into it. It was calculated from a self-report scale, the QIDS, rather than the Hamilton scale the protocol had specified. And it assumed that everyone who left the study early would have remitted at the same rate as those who stayed.10
A 2023 reanalysis of the patient-level data, following the original protocol, put the cumulative remission rate at 35.0 percent using the protocol’s Hamilton scale and its rules for who counted. The figure rose to 41.3 percent in a best-case version that added QIDS-defined remissions for patients missing a final Hamilton score.11
That reanalysis has its own assumptions, and it does not become the new one-number truth either. The honest summary is modest: later steps can still help, but the odds fall, more people drop out, and no sequence guarantees remission.
Urgency can change the order
Depression with psychosis, catatonia, trouble eating or drinking, strong suicidal intent, mania, or fast-dropping function should not wait behind a slow routine ladder. The safest next step may be an emergency assessment, hospital care, an ECT evaluation, or another fast, closely monitored treatment. A screening score cannot make that call by itself.
One-page appointment map
Bring one page with these headings:
- Treatments tried: name, highest dose, dates, how long, and whether you took it steadily.
- Benefit: which symptoms and which parts of daily life improved, how much, and for how long.
- Burden: side effects, cost, access, missed doses, and what made it hard to take.
- Therapy: type, number of sessions, fit, skills you use, and barriers.
- Diagnoses considered: unipolar depression, bipolar spectrum, trauma, psychosis, ADHD, grief, substance-related illness, or others your team raised.
- Sleep and substances: sleep schedule, snoring or apnea risk, alcohol, cannabis, stimulants, sedatives, and tobacco.
- Medical workup: relevant conditions, medication review, and tests already done.
- Safety now: suicidal thoughts, plan or intent, self-neglect, psychosis, mania, eating, drinking, housing, and support.
- Priorities: speed, avoiding weight or sexual side effects, memory, pregnancy, work, transport, cost, and visit burden.
- Access: insurance, distance, time off, pharmacy supply, language, and caregiver needs.
When to get help sooner
Get an urgent assessment for suicidal intent or an inability to stay safe. The same is true for psychosis, mania, catatonia, severe agitation, trouble eating or drinking, marked self-neglect, or function that is dropping fast. A new treatment plan can wait. Safety and medical needs come first.
If you may act on thoughts of suicide, or cannot stay safe, call or text 988 in the United States. Call 911 or go to the nearest emergency department for immediate danger.
What to do next
Fill in the one-page map. Then ask your clinician to name the branch you are on: diagnosis check, treatment optimization, switch, augmentation, or specialty referral. A named branch is more useful than another unexplained prescription.
Frequently asked questions
Does two failed antidepressants always mean TRD?
No. Many definitions use two adequate trials, but adequacy and even the diagnosis may need review.
Is augmentation better than switching?
Neither is always better. Partial benefit and tolerability may favor augmentation; no benefit or poor tolerance may favor switching.
Is ketamine the same as Spravato?
No. Spravato is FDA-approved intranasal esketamine with a restricted program. IV ketamine for depression is off-label.
Can a genetic test choose the right antidepressant?
No. Some results may inform dosing or interactions for certain drugs, but they cannot prove which treatment will work.
Is ECT only a last resort?
No. Severity, psychosis, catatonia, nutrition, suicide risk, prior response, and need for speed can move it earlier.
This article is for education, not personal medical advice. Do not start, stop, restart, or change a medication without the clinician who manages it. Every numeric, regulatory, and citation claim was verified against primary journal, federal, and labeling sources on September 7, 2026. Guidelines, drug labels, and public-health data change; confirm current status before acting on anything here.
Related reading on NP FADY
- IV ketamine for depression
- Spravato, or esketamine, for depression
- How TMS works
- What psychiatric genetic testing can and cannot tell you
- Low-dose naltrexone in psychiatry
- Modern ECT for Depression: Procedure, Benefits, Memory, and Safety
- The Two-Disease Loop: Depression With Diabetes or Heart Disease
References
1. McIntyre RS, Alsuwaidan M, Baune BT, et al. “Treatment-resistant depression: definition, prevalence, detection, management, and investigational interventions.” World Psychiatry. 2023;22(3):394–412. DOI: 10.1002/wps.21120. Open full text. See also Gaynes BN, Lux L, Gartlehner G, et al. “Defining treatment-resistant depression.” Depression and Anxiety. 2020;37(2):134–145. DOI: 10.1002/da.22968. PubMed record.
2. Lam RW, Kennedy SH, Adams C, et al. “Canadian Network for Mood and Anxiety Treatments (CANMAT) 2023 update on clinical guidelines for management of major depressive disorder in adults.” Canadian Journal of Psychiatry. 2024;69(9):641–687. DOI: 10.1177/07067437241245384. Open full text. Department of Veterans Affairs and Department of Defense. Clinical Practice Guideline for the Management of Major Depressive Disorder. February 2022. Guideline.
3. National Institute for Health and Care Excellence. Depression in adults: treatment and management (NG222). Published 29 June 2022; recommendations checked 2026. See recommendations 1.9.1 to 1.9.9. Further-line recommendations.
4. U.S. National Library of Medicine, DailyMed. Current manufacturer labels checked September 7, 2026: aripiprazole, Otsuka America Pharmaceutical, revised 1/2025, quetiapine XR, AstraZeneca Pharmaceuticals, revised 1/2025, and olanzapine-fluoxetine, Symbyax. For the point that olanzapine alone is not indicated for treatment-resistant depression, see the Zyprexa prescribing information, revised 1/2025. U.S. Food and Drug Administration. Current prescribing information: brexpiprazole, cariprazine, and lumateperone. Lumateperone was approved for adjunctive use in major depressive disorder on November 6, 2025. Lithium and liothyronine labels carry no major depressive disorder indication, which is the basis for describing those uses as off-label.
5. U.S. Food and Drug Administration. Repetitive Transcranial Magnetic Stimulation (rTMS) Systems - Class II Special Controls Guidance for Industry and FDA Staff. rTMS systems are Class II devices under 21 CFR 882.5805 and reach market through 510(k) clearance rather than premarket approval. FDA guidance.
6. U.S. Food and Drug Administration. Spravato (esketamine) prescribing information, supplement 019, revised 04/2025. FDA label. The monotherapy option for treatment-resistant depression was added by supplement 016, approved January 17, 2025.
7. Bousman CA, Stevenson JM, Ramsey LB, et al. “Clinical Pharmacogenetics Implementation Consortium (CPIC) guideline for CYP2D6, CYP2C19, CYP2B6, SLC6A4, and HTR2A genotypes and serotonin reuptake inhibitor antidepressants.” Clinical Pharmacology & Therapeutics. 2023;114(1):51–68. DOI: 10.1002/cpt.2903. Open full text. VA/DoD, February 2022, recommendation 13. Guideline.
8. Moloney BD, Forsyth A, Sumner RL, et al. “Low-dose naltrexone as an adjunctive treatment for major depressive disorder: findings from a randomized, double-blind, placebo-controlled hybrid parallel-arm study.” Frontiers in Pharmacology. 2026;1767654. DOI: 10.3389/fphar.2026.1767654. Thirty-seven patients randomized; MADRS scores fell 10.5 in the low-dose naltrexone group and 9.8 in the placebo group, with no difference between groups (p = 0.97). Verified at abstract level only. Open record.
9. National Institute of Mental Health. “STAR*D Study: All Medication Levels.” Study overview.
10. Rush AJ, Trivedi MH, Wisniewski SR, et al. “Acute and longer-term outcomes in depressed outpatients requiring one or several treatment steps: a STAR*D report.” American Journal of Psychiatry. 2006;163(11):1905–1917. DOI: 10.1176/ajp.2006.163.11.1905. This is the source of the 67 percent figure, which the paper itself describes as a theoretical cumulative remission rate that “assumes no dropouts” and assumes that those who exited would have remitted at the same rate as those who stayed. Journal record.
11. Pigott HE, Kim T, Xu C, Kirsch I, Amsterdam J. “What are the treatment remission, response and extent of improvement rates after up to four trials of antidepressant therapies in real-world depressed patients? A reanalysis of the STAR*D study’s patient-level data with fidelity to the original research protocol.” BMJ Open. 2023;13(7):e063095. DOI: 10.1136/bmjopen-2022-063095. Open full text.
If you or someone you know is in crisis
- Call 911 or go to your nearest emergency room for any life-threatening emergency.
- 988 Suicide & Crisis Lifeline — call or text 988, available 24/7. En español: marque 988 y oprima 2. Veterans: 988 and press 1, or text 838255.
- Crisis Text Line — text HOME to 741741.
- The Trevor Project (crisis support for LGBTQ+ young people) — call 1-866-488-7386, or text START to 678-678.
- National Sexual Assault Hotline (RAINN) — call 1-800-656-HOPE (4673) or text HOPE to 64673; free, confidential, 24/7. Online chat at RAINN.org/hotline.
- National Domestic Violence Hotline — call 1-800-799-SAFE (7233) or text START to 88788; 24/7, help in 200+ languages. Online chat at TheHotline.org. If your phone or computer may be monitored, calling from a safer device is an option.
- Riverside County — Inland SoCal Crisis Helpline 951-686-HELP (4357), 24/7 (Inland SoCal United Way / 211+, in partnership with RUHS-BH); Community Access, Referral, Evaluation and Support (CARES) Line 800-499-3008, 24/7.
- San Bernardino County — Access Unit (Behavioral Health Helpline) 888-743-1478, 24/7; Mobile Crisis/CCRT 800-398-0018 (24/7, all ages) or text 909-420-0560. Arrowhead Regional Medical Center (ARMC) has a dedicated walk-in adolescent psychiatric ER (ages 13–17).
- Children under 13 — call 911 for immediate danger, contact your county's mobile crisis team (they respond to all ages), or go to the nearest pediatric emergency room.
- California Peer-Run Warm Line (non-crisis — someone to talk to) — call or text 1-855-600-WARM (9276); daytime and evening hours, not a 24/7 line.
- NP Fady (non-emergency) — for routine scheduling or questions, call (909) 707-6261. This line is not monitored for emergencies.