Strategy for literature search

This review was conducted as a structured narrative review informed by PRISMA principles for literature identification and reporting.

In the first stage, we identified online repositories of biomedical literature, including Scopus, EMBASE, PubMed, and MEDLINE. Articles were retrieved from these databases using the following search strategy: ([ketamine] AND ((intravenous) OR (IV) OR (infusion)) AND [depression]). Duplicate records were automatically removed using the ZOTERO reference manager (PMID: 37684721).

Further screening of the retrieved records was performed according to predefined inclusion and exclusion criteria.

Inclusion criteria:

• – Full-text articles;

• – Published in peer-reviewed journals;

• – Published within the last 5 years;

• – Published in English;

• – Human studies;

• – Adults aged 19–65 years.

Exclusion criteria:

• – Duplicate records;

• – Diagnoses other than depression;

• – Non-human studies;

• – Lack of full-text availability in English;

• – Non-IV ketamine administration;

• – Combined treatments;

• – Case reports;

• – Studies involving children or adolescents;

• – Other non-relevant topics.

Additionally, a manual search was conducted to identify specific biomarkers mentioned in the text, yielding x additional publications.

Publications meeting all the above criteria underwent a full-text review. Figure 1 presents a schematic representation of the publication selection process.

Figure 1.

Flow diagram of a systematic selection of intravenous ketamine publications according to PRISMA guidelines (See also section “Strategy for literature search”). *Record excluded using automated tools built-in database and bibliography manager. **Record excluded by a researcher.

Figure 1. Long description

The flowchart is organized into three vertical phases: Identification, Screening, and Included.

1. Identification Phase:

- Records identified from databases: Scopus (n = 372), PubMed (n = 223), E M B A S E (n = 198), and M E D L I N E (n = 137).

- An arrow points right to Records removed before screening: Duplicate records (n = 290).

2. Screening Phase:

- Records screened (n = 639). An arrow points right to Records excluded: Other diagnoses (n = 147) and Non-Human (n = 6).

- Reports sought for retrieval (n = 486). An arrow points right to Reports not retrieved (n = 12).

- Reports assessed for eligibility (n = 474). An arrow points right to Reports excluded: Combined treatment (n = 221), Non intravenous ketamine (n = 59), Case reports (n = 23), Children and adolescents (n = 8), and Other (n = 13).

3. Included Phase:

- Studies included in review (n = 150).

Vertical arrows connect the central boxes from top to bottom, showing the reduction in study count at each stage.

Source: Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 2021;372:n71. doi:10.1136/bmj.n71 (http://www.prisma-statement.org/).

Inflammatory biomarkers

Elevated proinflammatory cytokines, including IL-6, TNF-α, and CRP, are associated with poor response to conventional antidepressants, particularly serotonergic and noradrenergic agents.Reference Zeng, Chourpiliadis and Hammar ²⁸ Ketamine’s immunomodulatory effects, such as reduced microglial activation and modulation of cytokine signaling, may enhance efficacy in patients with elevated inflammation.Reference Nikkheslat ^{29 ,} Reference Szałach, Ciesielska-Figlon, Daca, Cubała and Lisowska ³⁰ Recent evidence highlights baseline absolute monocyte count (AMC) as a stronger predictor of ketamine response than CRP, correlating with symptom improvement independent of age, sex, or BMI.Reference Pedraz-Petrozzi, Spangemacher and Deicher ³¹ Nevertheless, these findings originate from a limited number of studies and have not yet been independently replicated in large prospective cohorts.

Neuroplasticity biomarkers

BDNF is among the most frequently investigated candidate biomarkers of ketamine response and a pharmacodynamic marker of ketamine response; higher pre-treatment levels or post-infusion increases are linked to better clinical outcomes.Reference Zelada, Garrido and Liberona ^{32 ,} Reference Zolghadriha, Anjomshoaa, Jamshidi and Taherkhani ³³ However, results remain inconsistent across studies, and no consensus currently supports its clinical use as a predictive marker. Other neuroplasticity-related markers, including VEGF, neurofilament light chain, complement components, and selected cytokines, also correlate with clinical response, reflecting synaptic remodeling and CNS reparative processes.Reference Lin, Su and Li ^{34 –}Reference Zheng, Zhou and Wang ³⁸

Genetic and epigenetic predictors

Genetic variants modulate ketamine efficacy and durability. BDNF Val66Met polymorphism affects secretion and response dynamics; CYP2B6 variants influence pharmacokinetics; GRIN2B and other NMDA receptor gene variants are linked to efficacy and tolerability.Reference Chen, Lin and Tsai ^{19 ,} Reference Perez-Rodriguez, New and Goldstein ^{40 –}Reference Rodrigues, Mcintyre and Lipsitz ⁴³ GWAS suggest polygenic influences, though clinical application remains limited.Reference Bao, Zhao and Li ^{41 ,} Reference Rodrigues, Mcintyre and Lipsitz ⁴³ Epigenetic modifications, including DNA methylation and microRNA regulation of the HPA axis, neuroplasticity, and immune genes, further modulate ketamine sensitivity,Reference Zeng, Chourpiliadis and Hammar ^{28 ,} Reference Cathomas, Bevilacqua and Ramakrishnan ^{44 ,} Reference Georgiou, Farmer and Medeiros ⁴⁵ (Table 1). Despite growing interest in genetic and epigenetic predictors, available findings are characterized by small sample sizes, inconsistent replication, and limited predictive performance. Consequently, genetic testing is not currently recommended for routine clinical decisions.

Table 1.

Biomarkers to Intravenous Ketamine in TRD

Table 1. Long description

The table consists of five columns and three data rows.

Row 1: Inflammatory biomarkers.

- Examples: I L-6, T N F-alpha, A M C.

- Association: Higher A M C is associated with better antidepressant response, while C R P is inconsistent.

- Evidence: Prospective observational studies.

- Clinical readiness: Preliminary evidence.

Row 2: Neuroplasticity biomarkers.

- Examples: B D N F, V E G F, N f L, complement system components.

- Association: Higher baseline or post-infusion B D N F correlates with better response; V E G F and N f L changes reflect synaptic plasticity.

- Evidence: R C Ts and observational studies.

- Clinical readiness: Preliminary evidence.

Row 3: Genetic/Epigenetic biomarkers.

- Examples: B D N F Val 6 6 Met, C Y P 2 B 6, G R I N 2 B.

- Association: Genetic variants influence response and tolerability; D N A methylation and micro R N A expression affect sensitivity.

- Evidence: G W A S and candidate gene studies.

- Clinical readiness: Research use only.

Characteristics of depression

The clinical characteristics of a depressive episode represent one of the most well-documented and consistent predictors of response to intravenous ketamine. Numerous studies have demonstrated that specific symptom domains—particularly the severity of anhedonia, presence of suicidal ideation, and comorbid anxiety symptoms—significantly influence both the likelihood and temporal dynamics of response.Reference Oliver, Snyder, Feinn, Malov, McDiarmid and Arias ⁴⁷ Ketamine exhibits particularly rapid and robust anti-anhedonic and anti-suicidal effects, which clearly differentiate it from conventional antidepressants with delayed onset and limited efficacy in these domains.Reference Abdallah, Averill and Gueorguieva ^{48 ,} Reference McIntyre, Rodrigues and Lipsitz ⁴⁹

The depressive symptom profile qualitatively modulates treatment efficacy. Patients with non-melancholic, atypical, or anxiety-predominant depression may show a different, often faster response to ketamine than those with classical melancholic depression.Reference Papakostas ⁵⁰ These differences likely reflect distinct neurobiological mechanisms, including the degree of stress axis involvement, glutamatergic system engagement, and networks regulating emotion and reward.Reference Parikh, Lopez and Vande Voort ⁵¹ At the same time, severe anxiety symptoms have in some studies been associated with slower or less durable responses, highlighting population heterogeneity and potential moderating factors such as autonomic arousal or comorbid anxiety disorders.Reference Zheng, Zhou and Wang ^{38 ,} Reference Rodrigues, Mcintyre and Lipsitz ^{43 ,} Reference Park, Luckenbaugh and Pennybaker ^{52 –}Reference Zheng, Yang and Gu ⁵⁴

The presence of psychotic features has received relatively limited investigation. Although psychotic depression is not considered an absolute contraindication to intravenous ketamine administered under appropriate psychiatric supervision, available evidence is insufficient to determine whether psychotic symptoms independently influence treatment response. Additional prospective studies are warranted.Reference Zhou, Wang and Lan ⁵⁵

In BD, disease course has particular prognostic significance, including history of manic or mixed episodes and current mood stabilization. Ketamine can be effective in treating depressive episodes in BD, but its use carries the risk of mood destabilization, including induction of hypomanic or manic symptoms. Clinical data indicate that patients with Bipolar I disorder, particularly those with a history of rapid cycling or previous manic episodes, should remain on mood stabilizers before initiation of intravenous ketamine treatment in order to minimize the risk of affective switching. This recommendation is primarily based on expert consensus and clinical practice guidelines rather than robust randomized controlled trial evidence.Reference Zheng, Gu and Yang ^{37 ,} Reference Abdallah, Averill and Gueorguieva ^{48 ,} Reference Abdallah, Ahn and Averill ^{56 ,} Reference Liu, Zhou and Wang ⁵⁷

Sleep disturbances

They are common in TRD; up to 94% report significant sleep problems, and over 60% present more than one sleep phenotype, including hypersomnia, early insomnia, and REM/SWS architecture changes.Reference Liu, Zhou and Wang ^{57 ,} Reference Palagini, Geoffroy, Miniati, Riemann, Gemignani and Marazziti ⁵⁸ Clinical data indicate that patients with hypersomnia and atypical depression features show greater and faster symptom reduction after ketamine compared to those with melancholic depression, suggesting a link between neurovegetative depression profile and glutamatergic.Reference Zheng, Zhou and Wang ^{38 ,} Reference Rodrigues, Mcintyre and Lipsitz ^{43 ,} Reference Park, Luckenbaugh and Pennybaker ^{52 –}Reference Zheng, Yang and Gu ^{54 ,} Reference Patarroyo-Rodriguez, Pazdernik, Vande Voort, Kung and Singh ⁵⁹ Mechanistically, ketamine increases glutamate release in the prefrontal cortex and may enhance neuroplasticity in areas regulating sleep and mood, potentially explaining higher efficacy in certain depression phenotypes.Reference Sanacora, Frye and McDonald ^{2 ,} Reference Duman, Sanacora and Krystal ⁷ Objective sleep measures further support these clinical observations. Several polysomnographic studies have demonstrated that ketamine administration increases slow-wave sleep (SWS), enhances slow-wave activity, and modulates REM sleep architecture. These changes correlate with synaptic plasticity and antidepressant response.Reference Kwaśny, Cubała, Włodarczyk and Pastuszak ⁶⁰ Because patients with atypical depression and hypersomnia often exhibit disturbances in SWS regulation, these findings provide a plausible neurobiological explanation for their greater responsiveness to intravenous ketamine. Nevertheless, the number of polysomnographic studies remains limited, and larger prospective investigations are required before PSG markers can be considered clinically useful predictors of treatment outcome.Reference Borentain, Williamson and Turkoz ⁶¹

Degree of treatment resistance

Degree of TRD is another strong predictor.Reference Zheng, Zhou and Wang ^{38 ,} Reference Rodrigues, Mcintyre and Lipsitz ^{43 ,} Reference Park, Luckenbaugh and Pennybaker ^{52 –}Reference Zheng, Yang and Gu ⁵⁴ Patients with less advanced treatment resistance or fewer previously failed evidence-based treatments generally achieve better outcomes than those with more advanced treatment resistance, prolonged psychiatric hospitalization, or multiple previous treatment failures.Reference Lima Constantino, Godschalk and van Dalfsen ^{62 ,} Reference Pastre, Chancel, Malestroit, Courtet and Olié ⁶³ Previous treatment failures, including electroconvulsive therapy (ECT), may reflect a greater degree of illness refractoriness. However, current evidence remains insufficient to conclude that prior ECT nonresponse independently predicts reduced response to intravenous ketamine. Importantly, available studies rarely distinguish between lifetime exposure to ECT and failure of ECT during the current depressive episode, limiting interpretation of its prognostic value. Therefore, previous ECT nonresponse should not currently be considered an independent contraindication to intravenous ketamine treatment, although further prospective studies are needed.Reference Abdallah, Fasula, Kelmendi, Sanacora and Ostroff ^{64 ,} Reference Baldinger-Melich, Spies, Bozic, Kasper, Rujescu and Frey ⁶⁵ Comorbid anxiety symptoms may increase the likelihood of favorable response, although findings are inconsistent and may depend on methodology and study population.Reference Medeiros, Demo, Goes, Zarate and Gould ^{46 ,} Reference McIntyre, Rodrigues and Lipsitz ⁴⁹ In BD, disease course has particular prognostic significance, including history of manic or mixed episodes and current mood stabilization. Ketamine can be effective in treating depressive episodes in BD, but its use carries the risk of mood destabilization, including induction of hypomanic or manic symptoms.Reference Di Vincenzo, Lipsitz and Rodrigues ⁶⁶ Clinical data indicate that adequate pharmacological stabilization prior to ketamine administration is a key safety factor and may also influence treatment efficacy.Reference Zheng, Gu and Yang ^{37 ,} Reference Abdallah, Averill and Gueorguieva ^{48 ,} Reference Abdallah, Ahn and Averill ^{56 ,} Reference Liu, Zhou and Wang ^{57 ,} Reference Badathala, Roach, Pennybaker, Wallace, Mathalon and Marton ⁶⁷

Family history and substance use

One well-documented predictor is a positive first-degree family history of alcohol use disorder (AUD).Reference Phelps, Brutsche, Moral, Luckenbaugh, Manji and Zarate ⁶⁸ Studies showed that TRD patients with positive family AUD history responded more strongly and rapidly to ketamine within hours.Reference Phelps, Brutsche, Moral, Luckenbaugh, Manji and Zarate ^{68 ,} Reference Bonaventura, Lam and Carlton ⁶⁹ Subsequent meta-analyses confirmed that positive AUD family history is associated with significantly better response at both 1 and 7 days post-infusion.Reference Zhou, Liu and Zheng ¹⁴ Personal substance use history, especially alcohol, may also influence ketamine efficacy, though data are mixed. Some studies suggest greater response in those with prior AUD, while others show no significant difference, potentially related to abstinence duration, addiction severity, or psychosocial factors.Reference Phelps, Brutsche, Moral, Luckenbaugh, Manji and Zarate ^{68 ,} Reference Raquib, Nguyen and Brown ⁷⁰ Childhood trauma, higher anhedonia severity, and lack of melancholic features were also associated with better response in some studies, though contradictory findings exist.Reference Medeiros, Demo, Goes, Zarate and Gould ^{46 ,} Reference Park, Luckenbaugh and Pennybaker ^{52 ,} Reference Jonovska, Sugnet and Šendula-Jengić ⁷¹

Metabolic parameters

Metabolic parameters, including body mass index (BMI) and vitamin B12 levels, may modulate response to ketamine, although data remain partially inconsistent.Reference Szarmach, Cubała, Włodarczyk and Gałuszko-Węgielnik ²⁷ In some studies, higher BMI was associated with better therapeutic response, potentially due to effects on ketamine pharmacokinetics or low-grade inflammation, while other analyses did not confirm this relationship.Reference Szarmach, Cubała, Włodarczyk and Gałuszko-Węgielnik ^{27 ,} Reference Dale, Bryant and Thompson ^{72 –}Reference Sakurai, Setoyama and Kato ⁷⁴These discrepancies may result from differences in dosing regimens, response assessment methods, and heterogeneity of study populations.Reference Dale, Bryant and Thompson ⁷²

Vitamin B12 plays a critical role in central nervous system function, supporting DNA methylation, neurotransmitter synthesis, and myelin integrity.Reference Dean, Hurducas and Hawton ⁷⁵ B12 deficiency has been repeatedly linked to poorer antidepressant response, whereas adequate levels or supplementation may enhance the efficacy of biological therapies, including ketamine.Reference Malewska, Permoda-Osip and Bartkowska-Śniatkowska ^{73 ,} Reference Fava, Borus, Alpert, Nierenberg, Rosenbaum and Bottiglieri ^{76 ,} Reference Joseph, Parsaik, Ahmed, Erwin and Singh ⁷⁷

Biological sex, age, and concomitant pharmacotherapy

The influence of biological sex on ketamine efficacy remains inconclusive. Most clinical studies do not report significant differences in therapeutic response between females and males in terms of either magnitude or duration of antidepressant effect. Nonetheless, some reports have suggested slightly better response in males within the first 24 hours post-infusion, possibly related to hormonal, pharmacokinetic, or receptor sensitivity differences.Reference Malewska, Permoda-Osip and Bartkowska-Śniatkowska ^{73 ,} Reference Rybakowski, Permoda-Osip and Bartkowska-Sniatkowska ⁷⁸

Concomitant use of other psychotropic medications is an important factor modulating ketamine efficacy. Attention has focused particularly on benzodiazepines and opioid receptor antagonists such as naltrexone. Some studies have shown attenuation of ketamine’s antidepressant effect with high-dose benzodiazepines, likely due to their inhibitory influence on synaptic plasticity and prefrontal cortex activation. Similarly, opioid receptor blockade has in some studies been associated with reduced ketamine efficacy, suggesting involvement of the opioid system in its mechanism of action.Reference Williams, Heifets and Blasey ⁷⁹

However, other studies have not confirmed a significant impact of concomitant pharmacotherapy on clinical response, indicating potential modulation by dose, timing, and individual biological differences.Reference Gałecki, Bliźniewska-Kowalska and Cubała ^{17 ,} Reference Medeiros, Demo, Goes, Zarate and Gould ⁴⁶ Clinically, these factors should be considered when planning ketamine treatment, although they do not always necessitate modification of existing pharmacotherapy.Reference Pennybaker, Roach and Fryer ⁸⁰

Age has been investigated as a potential moderator of ketamine response, although findings remain inconsistent. Most studies suggest comparable antidepressant efficacy across adult age groups, whereas older patients are substantially underrepresented in clinical trials.Reference Badathala, Roach, Pennybaker, Wallace, Mathalon and Marton ⁶⁷ Consequently, current evidence does not support age as an independent predictor of response, although pharmacokinetic differences and medical comorbidities should be considered when treating elderly patients.Reference Badathala, Roach, Pennybaker, Wallace, Mathalon and Marton ^{67 ,} Reference Pennybaker, Roach and Fryer ⁸⁰

Neuroimaging and neurophysiological factors

Neuroimaging and neurophysiological markers are increasingly recognized as predictors of intravenous ketamine efficacy in TRD, reflecting susceptibility to rapid antidepressant action and response durability.

fMRI studies show baseline functional connectivity of the prefrontal cortex with the cingulate gyrus, hippocampus, and amygdala predicts ketamine response.Reference Nakamura, Tomita and Horikawa ^{81 –}Reference Yonezawa, Tani and Shibukawa ^{83 ,} Reference Zhang, Wang and Lan ⁸⁴ Disrupted networks in TRD are linked to rumination, emotional dysregulation, and cognitive deficits.Reference Gärtner, Weigand and Scheidegger ^{85 ,} Reference Milak, Rashid and Dong ⁸⁶ Ketamine modulates glutamatergic transmission, normalizing connectivity in default mode, executive, and salience networks.Reference Abdallah, Ahn and Averill ^{56 ,} Reference Nakamura, Tomita and Horikawa ^{81 ,} Reference Chen, Lin and Wu ⁸⁷

EEG and MEG markers, including gamma, theta, and alpha power, as well as interregional synchrony, predict ketamine response, with gamma activity reflecting glutamatergic activation and synaptic plasticity.Reference Chan, Olbrich, Jiang, Zhang, Ip and Brunovsky ^{88 –}Reference Kopelman, Keller and Panny ^{90 ,} Reference Zhang, Huang, Xu, Di, Yu and Zeng ⁹¹ MRS shows prefrontal glutamate/GABA ratios and their modulation by ketamine correlate with clinical improvement.Reference Chen, Lin and Tsai ^{19 ,} Reference Nakamura, Tomita and Horikawa ^{81 ,} Reference Chen, Lin and Wu ⁸⁷

Pilot protocols in ECT-nonresponsive patients use biweekly ketamine infusions with TMS-EMG–EEG assessments. Measures like intracortical facilitation (ICF) and short-interval cortical inhibition (SICI) reflect glutamatergic/GABAergic balance and ketamine’s mechanism.Reference Franklin, Altinay and Bailey ^{92 –}Reference Zavaliangos-Petropulu, McClintock and Khalil ⁹⁷

Resting-state EEG assessing vigilance and attention stability, eg, phase A1 proportion in cyclic alternating pattern (CAP) analysis, correlates with better and more stable ketamine response.Reference Monn, Eicher and Rüesch ^{98 ,} Reference Singh, McMahon and Bilderbeck ⁹⁹ Although neuroimaging and neurophysiological measures provide valuable mechanistic insights, their clinical applicability remains limited by high costs, methodological heterogeneity, lack of standardization, and absence of prospective validation studies.

Patient selection

Key to ketamine therapy success is appropriate patient selection, considering response predictors. Several studies suggest that certain patient characteristics may be associated with a higher probability of response, although findings remain inconsistent:

• • Atypical depression phenotype, including hypersomnia-type sleep disturbances,Reference Wang, Zhang and Zhou ¹¹² demonstrating higher glutamatergic sensitivity.Reference Niciu, Luckenbaugh and Ionescu ^{113 ,} Reference Reif-Leonhard, Millard and Ferdowssian ¹¹⁴

• • Moderate treatment resistance, without a long psychiatric hospitalization history or multiple suicide attempts .Reference Phillips, Norris and Talbot ^{115 ,} Reference Rengasamy, Mathew, Howland, Griffo, Panny and Price ¹¹⁶

• • Positive first-degree family history of AUD, correlating with faster and stronger response.¹¹⁷⁶⁸

• • Biological factors may indicate greater ketamine sensitivity and serve as predictive indicators.Reference Pedraz-Petrozzi, Spangemacher and Deicher ³¹

Dosing protocols and response monitoring

The most commonly used and well-documented protocol is a 0.5 mg/kg IV infusion over 40 minutes. Clinical practice indicates that repeated infusion series (typically 4–6 infusions over 2–3 weeks) can prolong therapeutic effect and increase response likelihood in TRD patients.Reference Sanacora, Frye and McDonald ^{2 ,} Reference Duman, Sanacora and Krystal ^{7 ,} Reference Gałecki, Bliźniewska-Kowalska and Cubała ¹⁷

Early response assessment is critical. After the first or second infusion, there is ≥20% reduction in depression scores (QIDS-SR16), which is a strong predictor of further response.Reference Lipsitz, McIntyre and Rodrigues ¹⁸ However, lack of early response does not exclude clinical benefit later; completing the planned protocol before final efficacy assessment is recommended.Reference Gutierrez, Kang and Vazquez ^{118 ,} Reference Haikazian, McIntyre and Meshkat ¹¹⁹

Concomitant medications and metabolic factors

Concurrent medications, especially benzodiazepines and opioid antagonists, may modulate ketamine response. High doses may partially attenuate effect; therefore, minimizing use or monitoring influence is recommended.Reference Quintanilla, Medeiros and Greenstein ^{36 ,} Reference Medeiros, Demo, Goes, Zarate and Gould ^{46 ,} Reference Williams, Heifets and Blasey ^{79 ,} Reference Meisner, Li and Boyle ¹²⁰

Metabolic factors, including vitamin B12, BMI, and childhood trauma history, may also influence efficacy. In cases of vitamin B12 deficiency or other adverse metabolic parameters, supplementation or lifestyle modification should be considered to improve treatment response p. 1.Reference Fava, Borus, Alpert, Nierenberg, Rosenbaum and Bottiglieri ⁷⁶

Integration of biomarkers and neurophysiology

Modern clinical approaches increasingly incorporate biomarkers and neurophysiology in monitoring ketamine response.Reference Wang, Lan and Liu ¹²¹ TMS combined with EEG and EMG assesses cortical inhibition-excitation balance (SICI/ICF), serving as a promising predictive biomarker.Reference Franklin, Altinay and Bailey ^{92 ,} Reference Zhang, Tan and Chen ¹²² Monitoring biomarkers such as BDNF or inflammatory markers remains investigational and should currently be considered a research tool rather than a standard component of clinical ketamine treatment.Reference Pedraz-Petrozzi, Spangemacher and Deicher ^{31 ,} Reference Wang, Zhang and Zhou ^{112 ,} Reference Zhang, Tan and Chen ¹²² This type of multimodal approach can significantly support therapy personalization and improve the effectiveness of interventions in TRD.

Specifics of bipolar disorder

In patients with depression in the course of bipolar affective disorder, it is important to ensure mood stabilization using mood-stabilizing drugs to minimize the risk of switching into a manic or hypomanic episode.Reference Sakurai, Setoyama and Kato ^{74 ,} Reference Sandström, Kampman and Asellus ¹²³ Studies indicate that ketamine shows comparable efficacy in depressive episodes of BD and MDD, making it a safe therapeutic option in severe bipolar depression when adequate mood-stabilizing prophylaxis is maintained.Reference Krystal, Abdallah, Sanacora, Charney and Duman ^{1 ,} Reference Sanacora, Frye and McDonald ²Based on current data, the following practical clinical implications can be formulated: appropriate patient selection considering depression phenotype, degree of treatment resistance, family history of AUD, and biological markers; use of repeated ketamine infusions with monitoring of early response; control of concomitant medications and metabolic parameters; use of neuroplasticity and neurophysiological biomarkers to predict response; and, in the case of BD, ensuring concurrent mood stabilization.Reference Price, Kissel and Baumeister ^{124 –}Reference Price, Wallace, Mathew and Howland ¹²⁶ This approach may help optimize treatment outcomes, minimizing the risk of adverse events, and supports the development of personalized psychiatry.Reference Tan, Gu and Zhou ^{107 ,} Reference Taraku, Woods and Boucher ¹²⁷

Integration of clinical, metabolic, inflammatory, and neuroplastic biomarkers, as well as depression phenotypes, may facilitate personalized treatment approaches. In clinical practice, this means the possibility of:

• • selecting patients with a high probability of response,

• • adjusting the infusion protocol,

• • monitoring early indicators of efficacy,

• • minimizing adverse effects,

• • developing maintenance treatment strategies and repeated infusions in long-term TRD therapy.Reference Tavares, da Costa Gonçalves and de Morais Barros ¹⁰⁹

A multimodal approach increases the safety, efficacy, and efficiency of treatment, and further research on neurophysiological and molecular markers of response may allow full implementation of a personalized treatment strategy for TRD in both MDD and BD,Reference Reif-Leonhard, Millard and Ferdowssian ¹¹⁴ (Figure 2).

Figure 2.

Predictors of IV ketamine response in TRD.

Figure 2. Long description

At the center of the diagram is a black circle with white text reading I V Ketamine Response – T R D. Eight lines radiate outward to rectangular nodes.

* Top-Left: Inflammatory / Immunological. Includes Absolute Monocyte Count (A M C) and C-Reactive Protein (C R P). Subtext: Routine Blood Tests.

* Top-Center: Metabolic. Includes Vitamin B 12 and Body Mass Index (B M I). Subtext: Standard Measurements.

* Top-Right: Clinical Phenotype. Includes Atypical Depression, Moderate Anxiety, and Suicidal Thoughts. Subtext: Clinical Assessment.

* Middle-Right: Treatment Resistance. Includes Failed Antidepressant Trials and Prior E C T / r T M S. Subtext: Medical History.

* Bottom-Right: Early Response Monitoring. Includes greater than or equal to 20% Improvement After 1–2 Infusions. Subtext: Rating Scales.

* Bottom-Center: Family / Substance History. Includes Alcohol Use Disorder (A U D). Subtext: Patient Interview.

* Bottom-Left: Family / Substance History. Includes Benzodiazepines and Opioid Antagonists. Subtext: Current Medications.

* Middle-Left: Concomitant Medications. Includes Benzodiazepines and Opioid Antagonists. Subtext: Current Medications.

Pilot clinical questionnaire for IV ketamine response (TRD in MDD and BD)

We developed a pilot clinical questionnaire to integrate predictors of ketamine response; see Supplementary Material for full details (Figure 3). The proposed questionnaire should be regarded as a hypothesis-generating research instrument and has not undergone prospective validation. It is not intended for routine clinical use.

Figure 3.

Clinical algorithm for IV ketamine candidate selection.

Relationship between IV ketamine and intranasal esketamine

Although this review focused on intravenous ketamine, the increasing clinical use of intranasal esketamine warrants a brief discussion.Reference Meisner, Li and Boyle ¹²⁰ Both interventions act on glutamatergic neurotransmission and demonstrate rapid antidepressant effects in patients with TRD. Intravenous ketamine typically leads to symptom improvement within hours of administration, whereas intranasal esketamine shows a more gradual therapeutic effect across repeated doses during the induction phase. In both cases, maintaining clinical benefit generally requires maintenance treatment; however, esketamine is used within more standardized regulatory frameworks.

Current data suggest that both treatment modalities may share similar predictors of response, such as early symptom improvement, inflammatory markers, and depression phenotype.Reference Chiappini, Guirguis and Schifano ¹⁴⁴ However, there is still a lack of head-to-head studies directly comparing the two methods, so it remains unknown whether non-response to intranasal esketamine is associated with a lower likelihood of response to intravenous ketamine. Moreover, there are currently no validated biomarkers available to guide the selection of the optimal therapy, highlighting the need for direct comparative studies of both interventions and further translational research.Reference Medeiros, Demo, Goes, Zarate and Gould ⁴⁶