A systematic review and meta-analysis of randomized controlled trials of adjunctive ketamine in electroconvulsive therapy: Efficacy and tolerability

Journal of Psychiatric Research 62 (2015) 23e30

Contents lists available at ScienceDirect

Journal of Psychiatric Research journal homepage: www.elsevier.com/locate/psychires

Review

A systematic review and meta-analysis of randomized controlled trials of adjunctive ketamine in electroconvulsive therapy: Efficacy and tolerability Alexander McGirr a, *, Marcelo T. Berlim b, c, David J. Bond d, Nicholas H. Neufeld e, Peter Y. Chan a, f, Lakshmi N. Yatham g, Raymond W. Lam a, g a

Department of Psychiatry, University of British Columbia, Vancouver, BC, Canada Neuromodulation Research Clinic, Douglas Mental Health University Institute and McGill University, Montr eal, Qu ebec, Canada Depressive Disorders Program, Douglas Mental Health University Institute and McGill University, Montr eal, Qu ebec, Canada d Department of Psychiatry, University of Minnesota, Minneapolis, MN, USA e Department of Psychiatry, University of Toronto, Toronto, ON, Canada f Neurostimulation Service, Vancouver General Hospital, Vancouver, BC, Canada g Mood Disorders Centre of Excellence, University of British Columbia, Vancouver, BC, Canada b c

a r t i c l e i n f o

a b s t r a c t

Article history: Received 25 September 2014 Received in revised form 7 January 2015 Accepted 8 January 2015

Background: Electroconvulsive therapy (ECT) remains one of the most effective tools in the psychiatric treatment armamentarium, particularly for refractory depression. Yet, there remains a subset of patients who do not respond to ECT or for whom clinically adequate seizures cannot be elicited, for whom ketamine has emerged as a putative augmentation agent. Methods: We searched EMBASE, PsycINFO, CENTRAL, and MEDLINE from 1962 to April 2014 to identify randomized controlled trials evaluating ketamine in ECT (PROSPERO #CRD42014009035). Clinical remission, response, and change in depressive symptom scores were extracted by two independent raters. Adverse events were recorded. Drop-outs were assessed as a proxy for acceptability. Metaanalyses employed a random effects model. Results: Data were synthesized from 5 RCTs, representing a total of 182 patients with major depressive episodes (n ¼ 165 Major Depressive Disorder, n ¼ 17 Bipolar Disorder). ECT with ketamine augmentation was not associated with higher rates of clinical remission (Risk Difference (RD) ¼ 0.00; 95%CI ¼ 0.08 to 0.10), response (RD ¼ 0.01; 95%CI ¼ 0.11 to 0.08), or improvements in depressive symptoms (SMD ¼ 0.38; 95%CI ¼ 0.41 to 1.17). Ketamine augmentation was associated with higher rates of confusion/ disorientation/prolonged delirium (OR ¼ 6.59, 95%CI: 1.28e33.82, NNH ¼ 3), but not agitation, hypertension or affective switches. Conclusion: Our meta-analysis of randomized controlled trials of ketamine augmentation in the ECT setting suggests a lack of clinical efficacy, and an increased likelihood of confusion. Individuals for whom adequate seizures or therapeutic response cannot be obtained have not been studied using randomized controlled designs. Additional research is required to address the role of ketamine in this population. © 2015 Elsevier Ltd. All rights reserved.

Keywords: Ketamine Ketofol Electroconvulsive therapy ECT Major depression

1. Background Electroconvulsive therapy (ECT) remains one of the most effective tools in the psychiatric treatment armamentarium, particularly for refractory depression. Indeed, rates of response as

* Corresponding author. 11th Floor, 2775 Laurel Street, Vancouver, BC V5Z 1M9, Canada. E-mail address: [email protected] (A. McGirr). http://dx.doi.org/10.1016/j.jpsychires.2015.01.003 0022-3956/© 2015 Elsevier Ltd. All rights reserved.

high as 90% are reported in major depressive disorder (Petrides et al., 2001), and even in the context of treatment resistant depression up to 60% of patients achieve clinical response following ECT (Prudic et al., 1996). Nevertheless, there remains a subset of patients who do not respond to ECT, or for whom clinically adequate seizures cannot be elicited. This population has spurred substantial research on stimulation parameters (Krystal and Weiner, 1994) as well as optimal electrode placement (Kellner et al., 2010). Even with treatment optimization, a substantial portion of patients show limited or

24

A. McGirr et al. / Journal of Psychiatric Research 62 (2015) 23e30

partial response. This has in turn stimulated interest in augmentation strategies in order to improve clinical response to ECT. Several augmentation strategies have been employed (Loo et al., 2010), including hyperventilation, caffeine, and remifentanil. More recently, ketamine has become a focus of research and attention in the ECT setting and recently systematically reviewed (Fond et al., 2014). Ketamine is an antagonist of the N-methyl-Daspartate (NMDA) receptor, an ionotropic subpopulation of glutamate receptors. At high doses it acts as a dissociative anesthetic, but has antidepressant effects with lower doses (McGirr et al., in press). It is unclear whether adjunctive ketamine improves rates of remission and response when used in conjunction with ECT. Ketamine is an attractive adjunctive agent in the ECT setting given that it is an anesthetic agent with limited anticonvulsant properties (Krystal et al., 2003). Indeed, adjunctive ketamine was associated with early case reports of effectiveness in patients receiving ECT (Ostroff et al., 2005). This was followed by open-label trials suggestive of increased effectiveness in the initial ECT sessions, but no benefit compared to treatment as usual after six sessions (Okamoto et al., 2010). Ketamine has also garnered attention for its putative ability to temper the cognitive side-effects associated with ECT, and has been associated with improvements in time to re-orientation (Krystal et al., 2003) and word recall (McDaniel et al., 2006). We performed a systematic review and meta-analysis of randomized, double-blind, placebo-controlled trials to assess the efficacy of ketamine as an adjunctive agent to ECT. In order to maximize the clinical relevance of our findings, we focused on clinical remission and response as outcomes, but we also examined changes in clinician-rated depression scores. Adverse events were recorded in order to determine safety and tolerability, while acceptability was assessed using drop outs as a proxy measure.

cases where potentially eligible studies were missing key data, their corresponding authors were contacted at least twice by e-mail at 2week intervals. Additional data was provided by the corresponding authors of all trials, with one exception (Wang et al., 2012).

2. Methodology of the literature review

2.4. Data synthesis and analyses

2.1. Search strategy

Analyses were performed using Comprehensive Meta-Analyses Version 2.0 (Biostat, Englewood, NJ, USA). Given that true treatment effects were likely to vary between studies given different methodological characteristics including patient selection, ketamine doses, anesthetic agents and electrode placements, we used a random-effects model (Riley et al., 2011). Intention-to-treat data were analyzed (Fergusson et al., 2002). The efficacy of ketamine, as well as its acceptability, was investigated by Risk Difference (RD) as well as Odds Ratio (OR) and the Number Needed to Treat (NNT) or Number Needed to Harm (NNH). We utilized Standardized Mean Differences (SMD) to quantify changes in depression scores pre- and post-treatment. With respect to SMDs, as we could not retrieve the correlations between pre- and post-ketamine measures from the individual RCTs we followed the recommendation of Rosenthal (Rosenthal, 1993) and assumed a conservative estimation of r ¼ 0.7. Heterogeneity was assessed using the Q statistics and I2 (Cooper et al., 2009) and two-tailed p-values reported. Values of p < 0.1 for the former and >35% for the latter were deemed as indicative of study heterogeneity (Borenstein et al., 2009). Finally, we used Funnel Plots and Egger's Regression Intercept (Egger et al., 1997) to test for the presence of publication bias (Borenstein et al., 2009; Cooper et al., 2009).

This protocol was registered in the PROSPERO registry (CRD42014009035). We identified articles for inclusion by searching MEDLINE, EMBASE, PsycINFO, and the Cochrane Central Register of Controlled Trials (CENTRAL) until April 15th, 2014. The search procedures (including syntaxes, parameters, and results) are described in detail in the Supplementary Material. Briefly, the search terms “ketamine”, “ketofol”, “electroconvulsive therapy” and “ECT” were utilized to identify randomized controlled trials. We also reviewed the bibliographies of published trials retained in this study for additional unidentified studies. 2.2. Study selection Studies were identified on the basis of their title, abstract and full text, and were included if they satisfied all of the following criteria (Higgins and Green, 2008): 1) Study Validity: Random allocation; double-blind (i.e., patients and clinical raters blinded to treatment conditions); controlled; parallel arm design; 5 subjects randomized per study arm; 2) Sample Characteristics: Subjects aged 18e75 years with a diagnosis of primary major depressive episode (unipolar or bipolar) according to DSM-IV(APA, 1994) or ICD (WHO, 1992) criteria; 3) Treatment Characteristics: Ketamine given as an adjunct to ECT; 4) Publication-Related: Articles written in English. Studies were excluded if they: 1) Enrolled subjects with “narrow” diagnoses (e.g., postpartum depression) or secondary depression (e.g., vascular depression); 2) Did not report raw data or the authors did not provide raw data. In

2.3. Data extraction Data were recorded by two independent observers with subsequent review and consensus in a structured fashion as follows: Sample Characteristics e Mean age, sex, and primary diagnosis ECT related e Electrode placement, number of sessions Ketamine related e Dose and administration Control condition e The control condition and its associated characteristics were recorded Primary Outcome Measure e Clinical Remission, defined as a Hamilton Depression Rating Scale [HRDS](Hamilton, 1960) score of 7 for the 17-item version, of 8 for the 21-item version, 9 for the 25-item version, or a score of 8 for the MontgomeryeAsberg Depression Rating Scale [MADRS](Montgomery and Asberg, 1979). These definitions were those employed in the RCTs at the conclusion of ECT treatment. Secondary Outcome Measures e Clinical response, defined as a 50% reduction in post-treatment scores based on the study's primary efficacy measure (HRDS or MADRS) at study end; Change in clinician-rated depressive symptoms pre- and postintervention were recorded, as was seizure duration and cognitive testing. Secondary outcomes were defined at the conclusion of ECT treatment. Acceptability and Tolerabilitye Adverse events and overall dropout rates at study end.

3. Results 3.1. Literature search Our literature search is detailed in Fig. 1 and the Supplementary material (SupplFigs. 1e4; STable1). Study quality was assessed

23.92 ± 3.78

27.02 ± 8.51

MDD e Treatment n ¼ 34; n ¼ 2 MADRS Resistant withdrawn MDD n ¼ 31 HDRS-21 Jarventausta et al., 2013 Yoosefi et al., 2014

32.39 ± 6.28 n ¼ 51; n ¼ 5 MADRS withdrawn Loo et al., 2012 MDD þ BD

24.90 ± 1.57 n ¼ 18; n ¼ 2 HDRS-25 withdrawn MDD þ BD Abdallah et al., 2012

Abbreviations: RCT ¼ Randomized Controlled Trial, MDD ¼ Major Depressive Disorder, BD ¼ Bipolar Disorder, HDRS ¼ Hamilton Depression Rating Scale, MADRS ¼ Montgomery-Asberg Depression Rating Scale, F ¼ Female, M ¼ Male.

Thymatron; Half-age method

6 ECTs

A) 48.8 13F/19M B) 53.7 43.82 ± 2.13 14F/17M 6 ECTs Thymatron; Titration method

Right Unilateral and Bitemporal Bitemporal

9.10 ± 4.34 ECTs 43.33 ± 1.95 28F/18M Mecta; Titration method Right Unilateral

47.15 ± 0.98 7F/9M 6 ECTs Mecta; Titration method Right Unilateral and Bitemporal

56.20 ± 2.06 18M/22F 1 ECT Huicheng; Half-age method Bitemporal n ¼ 48; n ¼ 8 HDRS-17 withdrawn MDD Wang et al., 2012

Sample size; withdrawals Table 1 Characteristics of included studies.

Five RCTs were included in our meta-analysis, totaling 182 subjects with a major depressive episode (n ¼ 165 with MDD; n ¼ 17 with BD; Table 1). Two studies had a mixed unipolar and bipolar depression sample: n ¼ 10 with MDD and n ¼ 8 with BD (Abdallah et al., 2012); n ¼ 42 with MDD and n ¼ 9 with BD (Loo et al., 2012). Only one RCT involved patients with treatmentresistant MDD (Jarventausta et al., 2013). No RCT focused on ECT non-responders, and no RCT compared ketamine augmentation to a different augmentation strategy. One of the included studies was stopped for futility at a planned interim analysis (Abdallah et al., 2012). In all studies, ketamine was administered as an intravenous bolus. Two studies used ketamine at a dose of 0.5 mg/kg (Abdallah et al., 2012; Loo et al., 2012), while the remaining studies used doses of 0.4 mg/kg (Jarventausta et al., 2013), 0.8 mg/kg (Wang et al., 2012) and 1e2 mg/kg (Yoosefi et al., 2014). Ketamine was compared to thiopental 2e3 mg/kg in one trial (Yoosefi et al., 2014), and used in conjunction with thiopental at doses of 3.5 mg/kg (Abdallah et al., 2012) and 3e5 mg/kg (Loo et al., 2012) in two trials. Propofol was reported at a loading dose of 1.5 mg/kg with no

Diagnosis

3.2. Included RCTs: main characteristics

Study

Instrument Depression Design score (M ± SE)

using the Cochrane Collaboration's Tool for Assessing Risk of Bias (Higgins et al., 2011) concomitantly to eligibility (STable2). We identified 7 RCTs, 5 of which met inclusion criteria for the present investigation (Abdallah et al., 2012; Jarventausta et al., 2013; Loo et al., 2012; Wang et al., 2012; Yoosefi et al., 2014). One RCT was excluded because it lacked a clinician administered depressive rating scale (Rasmussen et al., 2014). One RCT was excluded as it focused on anesthetic and physiological parameters, and the absence of psychiatric outcomes was confirmed in correspondence with the authors (Yalcin et al., 2012).

25

A) Ketamine (0.8 mg/kg) B) Ketamine (0.8 mg/kg) þ Propofol (1.5 mg/kg) C) Propofol (1.5 mg/kg) A) Ketamine (0.5 mg/kg) þ Thiopental (3.5 mg/kg) B) Thiopental (3.5 mg/kg) A) Thiopental (3e5 mg/kg) þ Ketamine (0.5 mg/kg) B) Thiopental (3e5 mg/kg) þ Saline A) S-Ketamine (0.4 mg/kg) þ Propofol B) Saline þ Propofol A) Ketamine (1e2 mg/kg) B) Thiopental (2e3 mg/kg)

Fig. 1. Study selection PRISMA Flowchart.

29.61 ± 2.21

Electrode placement ECT system; seizure threshold Number of ECTs

Age (M ± SE) Sex

A. McGirr et al. / Journal of Psychiatric Research 62 (2015) 23e30

26

A. McGirr et al. / Journal of Psychiatric Research 62 (2015) 23e30

Fig. 2. Meta-analysis of A) rates of clinical remission and B) rates of clinical response.

information after loading in one study (Wang et al., 2012) and loading dose of 0.5 mg/kg and mean total dose of 99.5 mg/kg in the other (Jarventausta et al., 2013). Ketamine was used alone as an induction agent in two studies (Wang et al., 2012; Yoosefi et al., 2014). Two trials used a saline placebo condition (Jarventausta et al., 2013; Loo et al., 2012). Electrode placements were bitemporal (Wang et al., 2012; Yoosefi et al., 2014), mixed bitemporal and right unilateral (Abdallah et al., 2012; Jarventausta et al., 2013), and right unilateral (Loo et al., 2012). The method for determining seizure threshold was an age method in two RCTs (Wang et al., 2012; Yoosefi et al., 2014) and dose-titration method in three RCTs (Abdallah et al., 2012; Jarventausta et al., 2013; Loo et al., 2012). In three RCTs, the protocol involved 6 ECT sessions (Abdallah et al., 2012; Jarventausta et al., 2013; Yoosefi et al., 2014), one RCT involved 9.10 ± 4.34 ECT sessions (Loo et al., 2012), while one RCT involved a single ECT session (Wang, 2013). Efficacy analyses were performed with and without the single ECT trial due to marked methodological differences. Demographic characteristics were as follows: mean age was 47.15 (SD 0.71) years with 43.0%male/56.9% female. Concomitant medications were only detailed in one trial (Loo et al., 2012). Disparate clinician rated depression instruments were used across studies, including the HDRS-17 (Wang et al., 2012) HDRS-21 (Yoosefi et al., 2014) HDRS-25 (Abdallah et al., 2012) and MADRS (Jarventausta et al., 2013; Loo et al., 2012).

outcome (Fig. 2A). Similarly, 27/68 (39.7%) ketamine treated and 30/66 (45.4%) of placebo treated patients were classified as clinical responders for a pooled RD of 0.01 (95%CI ¼ 0.11 to 0.08; p ¼ 0.81; Fig. 2B) and OR of 1.13 (95%CI: 0.21e2.67). It is important to note that one study reported no patients meeting criteria for clinical response (confirmed with raw data) in either treatment arm after 6 ECT sessions (Yoosefi et al., 2014), and after excluding this study overall higher rates of remission (31.3% ketamine vs 28.8% control, ns) and response (52.9% ketamine vs 57.6% control, ns) were observed. Overall, there was no evidence of heterogeneity between RCTs (clinical response Q ¼ 4.15, I2 ¼ 27.75, p ¼ 0.24). Nevertheless, we examined the potential impact of methodological differences between studies. There was no evidence of heterogeneity based on concomitant anesthetic (clinical remission Q ¼ 0.03, p ¼ 0.85; clinical response Q ¼ 2.25, p ¼ 0.13), electrode placement (clinical remission Q ¼ 0.22, p ¼ 0.89; clinical response Q ¼ 0.88, p ¼ 0.64) or method of determining seizure threshold (clinical remission Q ¼ 0.03, p ¼ 0.84; clinical response Q ¼ 0.09, p ¼ 0.75). Sensitivity analyses revealed that the lack of difference in outcomes favoring ketamine with respect to clinical remission was not attributable to any outlying RCT (RDs  j0.02j, Z  j0.23j, p  0.81). Similarly, lack of difference in outcomes with respect to clinical response was not attributable to any outlying RCT (RDs  j0.03j, Z  j0.72j, p  0.46).

5. Depressive symptoms 4. Remission and response rates Data relating to remission and response rates were available for 4 RCTs, all but the RCT by Wang and colleagues (Wang et al., 2012). With respect to clinical remission, 16/68 (23.5%) of ketaminetreated and 15/66 (22.7%) of control-treated patients achieved remission for a pooled RD of 0.00 (95%CI ¼ 0.08 to 0.10; p ¼ 0.89) and OR of 0.83 (95%CI: 0.23e2.92), indicating no difference in

Data relating to continuous scores for clinician administered depression rating scales were available for all five trials. Overall, a non-significant SMD of 0.38 (95%CI ¼ 0.41 to 1.17, p ¼ 0.34; Fig. 3) was observed, with significant heterogeneity driven by one study (Wang et al., 2012), the previously discussed study with multiple methodological differences including employing a single ECT session (Q ¼ 27.70, I2 ¼ 85.56, p < 0.001). When analyses were

A. McGirr et al. / Journal of Psychiatric Research 62 (2015) 23e30

27

Fig. 3. Meta-analysis of change in clinician administered depression rating scores.

repeated after removing this study, a SMD of 0.06 (95%CI ¼ 0.28 to 0.41, p ¼ 0.70) was observed without significant heterogeneity (Q ¼ 3.12, I2 ¼ 3.86, p ¼ 0.37). 6. Seizure duration Data relating to continuous seizure duration was available for all five trials. A significant difference in means was observed (8.17 s, 95%CI: 1.75e14.59, p < 0.05) favoring ketamine. There was significant heterogeneity (Q ¼ 43.85, I2 ¼ 90.87, p < 0.001), with no clear relationship to anesthetic agent or electrode placement. 7. Adverse events, acceptability and tolerability Adverse events were detailed in all five trials, however definitions varied. These included agitation, confusion/prolonged delirium, and hypertension (Fig. 4). Two affective switches were noted, one into hypomania and one into a mixed state, both among patients with lithium-treated bipolar depression receiving ketamine augmentation (Loo et al., 2012). An increased risk of confusion/disorientation/delirium was noted amongst ketamine treated patients (OR ¼ 6.59, 95%CI: 1.28e33.82, p < 0.05, NNH ¼ 3). Other adverse events did not reach significance (Hypertension OR ¼ 2.75, 95%CI: 0.80e9.28; Agitation OR ¼ 2.05, 95%CI: 0.67e6.30).

Dropouts were examined as a proxy for tolerability. Among ketamine treated patients, a dropout rate of 11.0% (11/100) compared to 9.7% (8/82) of placebo treated patients (RD ¼ 0.01, 95% CI: 0.09 to 0.12, p ¼ 0.85) without evidence of heterogeneity (Q ¼ 6.06, I2 ¼ 34.03, p ¼ 0.19). 8. Cognitive side effects of ECT Only two of the RCTs included measures of cognition. The first involved a neuropsychological battery administered by a trained neuropsychologist, which included the autobiographical memory interview. This trial found no benefit associated with ketamine (Loo et al., 2012). The second utilized the MMSE and found a statistically significant benefit in favor of ketamine, where a grouped mean difference of 0.81 at baseline grew to 2.08 following ECT (Yoosefi et al., 2014). 9. Publication bias With respect to clinical response and remission, the associated Funnel Plots revealed broadly symmetrical distributions (SupplFig. 5A&B), indicating a marginal risk of publication bias. The risk of publication bias was also assessed with Egger's regression intercept, which for clinical remission was 0.34 (df ¼ 1, t ¼ 0.53,

Fig. 4. Adverse events and side effects.

28

A. McGirr et al. / Journal of Psychiatric Research 62 (2015) 23e30

two-tailed p ¼ 0.68) and for response data was 0.30 (df ¼ 1; t ¼ 0.08; two-tailed p ¼ 0.94). These findings suggest low potential for publication bias. With respect to continuous depression scores, the associated Funnel Plots (SupplFig. 5C&D) revealed a significant left shift, however repeating the analysis excluding the Wang et al. study (Wang et al., 2012) revealed a symmetrical distribution. Egger's regression intercept was not significant (6.73, df ¼ 3, two-tailed t ¼ 0.38). This suggests low possibility of publication bias. With respect to confusion, we examined the Fail-Safe N, which revealed that 6 RCTs would be required to render the increased risk of confusion associated with ketamine non-significant. 10. Discussion To our knowledge, this is the first meta-analysis of ketamine as an adjunctive agent to electroconvulsive therapy to examine clinical efficacy with the gold standard of clinical remission and response. In a pooled sample of 182 depressed individuals with unipolar or bipolar depression from 5 randomized controlled trials, our analyses suggest a lack of efficacy as indicated by measures of clinical response, clinical remission, or change in severity of depressive symptoms associated with adjunctive ketamine use in the ECT setting, despite prolonging seizure duration. In conjunction with evidence for lack of efficacy, our analyses suggest an increased risk of confusion/disorientation/prolonged delirium following ketamine-augmented ECT with a NNH of 3, though the clinical relevance is unclear. Our results are in stark contrast to the literature on ketamine as a novel treatment for depression (McGirr et al., in press). Unlike other investigated agents such as nortriptyline (Sackeim et al., 2009), ketamine does not appear to be an efficacious adjunctive agent in ECT, a conclusion further supported by null results reported in the RCT that was excluded from the study due to the absence of a clinician administered depression rating scale (Rasmussen et al., 2014). However, it is important to note that most patients included in the ketamine augmentation trials were not identified as patients for whom adequate seizures could not be obtained. This question is of significant clinical importance, and requires rigorous randomized controlled trials to determine if adjunctive ketamine to ECT is efficacious in ECT non-responders and those whose ECT stimulus parameters cannot be increased despite inadequate seizures. The lack of efficacy from this meta-analysis is consistent with open-label data demonstrating early superiority of adjunctive ketamine with ECT, yet no evidence of benefit later in the index course (Okamoto et al., 2010). It is perhaps unsurprising, therefore, that the only RCT to report a significant reduction in depressive symptoms with adjunctive ketamine involved a single ECT session (Wang et al., 2012). Nevertheless, the issue of early improvement deserves consideration, for while this difference from placebo may diminish with continued ECT, the potential for temporizing relief should not be discounted, especially for those who are highly suicidal or medically compromised by poor intake. Nevertheless, this proposition has not borne out in RCTs, and the one RCT to show early improvement had very low effect sizes as evidenced by the absence of clinical responders in either treatment group (Yoosefi et al., 2014). The results from the single administration ECT study mirror those from the literature involving single administrations of ketamine (McGirr et al., in press), where ketamine, when administered as a single intravenous infusion (Berman et al., 2000; Diazgranados et al., 2010; Murrough et al., 2013; Sos et al., 2013; Zarate et al., 2012, 2006), and more recently intra-nasally (Lapidus et al.,

2014), has shown rapid antidepressant effects. Similarly, repeated ketamine infusions have been demonstrated to result in greater improvement than the same number of ECT sessions (Ghasemi et al., 2014). Yet, unlike RCTs employing single infusions of ketamine, continued improvement was observed over the subsequent 7 days after a single session of ECT with ketamine (Wang et al., 2012). This fostered optimism regarding the potential for synergistic action between ketamine and ECT, with the potential for decreased burden on neurostimulation services. However, this conclusion has been revised in light of post-publication correspondence revealing that the improvement was transient and that patients went on to require additional ECT during the same hospital stay (Kellner et al., 2013; Wang, 2013). While ketamine increased seizure length, an associated improvement in clinical outcomes was not observed, and may be in part explained by limited utility of extending seizures beyond 25 s (Krystal et al., 2000). Seizure adequacy as defined by duration, however, is an over simplification (Krystal and Weiner, 1994). Indeed, numerous seizure characteristics, including typical features on EEG such as polyspike activity and post-ictal suppression, are important indicators of seizure adequacy (Whitehouse and Scott, 2005). When compared to methohexital, ketamine is associated with greater mid-ictal low-frequency EEG amplitude and a greater degree of post-ictal suppression (Krystal et al., 2003). It is, therefore, unclear whether lack of clinical benefit reflects a ceiling effect, or an as of yet unidentified mechanism. The optimal dosing of ketamine for the treatment of depression is being actively investigated, and there is emerging evidence that very low dose is associated with a lower degree of efficacy (Lai et al., 2014; Lapidus et al., 2014). Though the majority of trials in our review used a dose range of 0.4e0.8 mg/kg, one study employed a significantly greater dose (1e2 mg/kg) (Yoosefi et al., 2014). Though the common characteristic amongst these trials is a failure of ketamine to separate from the comparator condition, there is currently insufficient evidence to address the issue of doseeresponse in the ECT setting. It is worth noting, however, that higher doses employed in RCTs for depression in surgical settings have employed doses at the upper limit of those employed in the ECT trials examined and found evidence of efficacy (Kudoh et al., 2002). Anesthetic agents may be an important confound in the ECT setting. Indeed, one hypothesized mechanism for ketamine's actions are inhibition of GABAergic neurons (Homayoun and Moghaddam, 2007), and therefore their potentiation using barbiturate anesthetics may counteract ketamine's antidepressant action. Moreover, direct inhibition of AMPA receptors by barbiturates (Kamiya et al., 1999) may further compound this effect, as AMPA activation has been implicated in the antidepressant effects of ketamine in animal models (Koike et al., 2011). Two human trials provide data that is not subject to these mechanistic considerations, with conflicting results. One found evidence for significant improvement in depressive symptoms within the methodological limitations of a single session of ECT (Wang et al., 2012), while the other did not find evidence of superiority after an index course (Yoosefi et al., 2014). Putative pro-cognitive characteristics of ketamine in the ECT setting have been hypothesized, bolstered by observations including an earlier time to reorientation (Krystal et al., 2003) and improved working memory (McDaniel et al., 2006). Two trials in this meta-analysis assessed the effects of ketamine on cognition in patients receiving ECT, with conflicting results. One study using the MMSE demonstrating a modest benefit favoring ketamine (Yoosefi et al., 2014), while the other failed to identify benefit despite an extensive neuropsychological battery including an autobiographical memory test (Loo et al., 2012). The studies that assessed the

A. McGirr et al. / Journal of Psychiatric Research 62 (2015) 23e30

effects of ketamine infusion in non-ECT treatment resistant depression subjects also failed to reveal any consistent evidence for pro-cognitive benefits with ketamine infusion (Murrough et al., 2014). Our analyses demonstrated an increased burden of adverse events, specifically with respect to confusion/disorientation/ delirium among ketamine treated patients. It is important to note, however, that the clinical significance of this outcome is unclear as a low threshold was used in two of three trials reporting this outcome (Jarventausta et al., 2013; Loo et al., 2012). In support of our association, Rasmussen et al. (2014) also report higher rates of confusion and a marginally higher rate of psychotomimetic effects in their RCT of ketamine augmentation to ECT. While the significance is unclear, it may be especially concerning in elderly patients undergoing ECT. Hypertension and agitation were the other adverse events reported in trials, however neither was found to be increased in ketamine treated patients. The number of affective switches with ketamine augmentation in bipolar disorder patients (11.7% vs 0.0%) was only marginally higher than the 8% incidence reported in the bipolar disorder ECT literature (Angst et al., 1992; Bailine et al., 2010). Interestingly, a recently described case series has highlighted not only the adverse effects, but also the poor acceptability to patients of ketamine augmentation in ECT (Rasmussen and Ritter, 2014). Our analyses highlight the potential harm from utilizing ketamine as an augmentation agent in the ECT setting, and that the decision to proceed with this strategy must take place with careful consideration of the potential risks and benefits. 11. Limitations A limitation levied against the meta-analytical method is the combination of heterogeneous studies, poor-quality or unrepresentative studies, or the potential of publication bias. Differences in study design was evident among the included RCTs, with differences in ketamine dose, concomitant anesthetic agents, and stimulation parameters including electrode placement, as well as different depressive symptom rating scales. Though we cannot definitively exclude the possibility of such influences, we have attempted to temper these by using a comprehensive systematic review of the literature, assessing the quality of studies, and by examining both publication bias and heterogeneity. Additional potential sources of bias deserve mentioning. The study designs may have employed inadequate index courses as evidenced by low remission rates and the fact that one RCT had no patients who could be classified as clinical responders (Yoosefi et al., 2014). Moreover, the only RCT identified through our systematic review to support the use of ketamine employed a single ECT session (Wang et al., 2012) and drove heterogeneity in depressive symptom analyses. One RCT was terminated prematurely due to futility (Abdallah et al., 2012), and a one RCT had a brief modification of concomitant anesthetic due to shortage (Loo et al., 2012). An important limitation pertains to the lack of treatment resistant patients in the studies. There has been interest in the differential efficacy of ketamine in the treatment of unipolar as compared to bipolar major depressive episodes (Fond et al., 2014; McGirr and Berlim, 2014; McGirr et al., in press), however the limited number of bipolar participants in the included trials precludes this comparison. Higher rates of confusion/disorientation/prolonged delirium may be confounded by the psychotomimetic effects of ketamine, yet higher rates of prolonged delirium were also noted among ketamine treated patients. Our meta-analysis remains limited by a moderate total sample size, and therefore the possibility of both Type I and Type II error cannot be excluded.

29

A major limitation is the applicability to patients for whom therapeutic seizures cannot be obtained in the absence of adjunctive agents. This is clearly an area deserving of additional research attention. An additional question deserving of investigation relates to the minimal therapeutic dose for therapeutic efficacy and the optimal dose, as does the most effective anesthetic combination, including using ketamine in isolation. Similarly, the long-term safety has not been addressed. 12. Conclusion Our meta-analysis of randomized controlled trials of ketamine in the ECT setting did not find evidence for clinical efficacy, in conjunction with an increased risk of confusion. The use of ketamine among patients for whom an adequate seizure or therapeutic response cannot be obtained has yet to be subjected to level I scrutiny. Additional research is required address the role of ketamine in this important population. Role of funding source There is no funding source to declare for this study. Contributors AM performed the literature review. AM&MTB screened and extracted data. AM analyzed the data. AM drafted the manuscript. MTB, DJB, NHN, PYC, LNY, & RWL critically revised the manuscript. All authors approved the final manuscript. Conflicts of interest & disclosures No sources of funding or industry involvement in this systematic review and meta-analysis. AM, MTB, DJB, NHN, PYC, LNY & RWL declare that they have no conflicts of interest for this manuscript. DJB has received speaking fees or acted as a consultant for: the Canadian Network for Mood and Anxiety Treatments (CANMAT), the Canadian Psychiatric Association, Pfizer, Sunovion, BMS, Otsuka, Astra-Zeneca, Janssen-Ortho and Myriad; and has received research support from: the Canadian Institutes of Health Research (CIHR), the UBC Institute of Mental Health/Coast Capital Depression Research Fund, and Pfizer. LNY has received research grants from or is on speaker/advisory boards for AstraZeneca, Bristol-Myers Squibb, Canadian Institutes of Health Research, Canadian Network for Mood and Anxiety Treatments, Eli Lilly & Co., GlaxoSmithKline, Janssen, Michael Smith Foundation for Health Research, Novartis, Pfizer, Ranbaxy, Servier, and the Stanley Foundation. RWL is on ad hoc Speaker/Advisory Boards for, or has received research funds from, AstraZeneca, Biovail, Bristol-Myers Squibb, Canadian Institutes of Health Research, Canadian Network for Mood and Anxiety Treatments, Canadian Psychiatric Association Foundation, Eli Lilly, Litebook Company, Lundbeck, Lundbeck Institute, Mochida, Pfizer, Servier, St. Jude Medical, Takeda, and UBC Institute of Mental Health/Coast Capital Savings. Acknowledgments We would like to thank Drs. Abdallah, Ghaeli, Jarventausta, Kampman, and Loo for providing data not contained within the published versions of their trials.

30

A. McGirr et al. / Journal of Psychiatric Research 62 (2015) 23e30

Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.jpsychires.2015.01.003.

References Abdallah CG, Fasula M, Kelmendi B, Sanacora G, Ostroff R. Rapid antidepressant effect of ketamine in the electroconvulsive therapy setting. J ECT 2012;28: 157e61. Angst J, Angst K, Baruffol I, Meinherz-Surbeck R. ECT-induced and drug-induced hypomania. Convuls Ther 1992;8:179e85. APA. Diagnostic and statistical manual of mental disorders (DSM-IV). 4th ed. Washington, DC: American Psychiatric Association; 1994. Bailine S, Fink M, Knapp R, Petrides G, Husain MM, Rasmussen K, et al. Electroconvulsive therapy is equally effective in unipolar and bipolar depression. Acta Psychiatr Scand 2010;121:431e6. Berman RM, Cappiello A, Anand A, Oren DA, Heninger GR, Charney DS, et al. Antidepressant effects of ketamine in depressed patients. Biol Psychiatry 2000;47: 351e4. Borenstein M, Hedges LV, Higgins JPT, Rothstein HR. Introduction to meta-analysis. West Sussex, England: Wiley & Sons Ltd; 2009. Cooper H, Hedges LV, Valentine JC. The handbook of research synthesis and metaanalysis. New York, U.S.: Russell Sage Foundation Publications; 2009. Diazgranados N, Ibrahim L, Brutsche NE, Newberg A, Kronstein P, Khalife S, et al. A randomized add-on trial of an N-methyl-D-aspartate antagonist in treatment-resistant bipolar depression. Arch Gen Psychiatry 2010;67:793e802. Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ 1997;315:629e34. Fergusson D, Aaron SD, Guyatt G, Hebert P. Post-randomisation exclusions: the intention to treat principle and excluding patients from analysis. BMJ 2002;325: 652e4. Fond G, Loundou A, Rabu C, Macgregor A, Lancon C, Brittner M, et al. Ketamine administration in depressive disorders: a systematic review and meta-analysis. Psychopharmacology 2014;20:20. Ghasemi M, Kazemi MH, Yoosefi A, Ghasemi A, Paragomi P, Amini H, et al. Rapid antidepressant effects of repeated doses of ketamine compared with electroconvulsive therapy in hospitalized patients with major depressive disorder. Psychiatry Res 2014;215:355e61. Hamilton M. A rating scale for depression. J Neurol Neurosurg Psychiatry 1960;23: 56e62. Higgins JP, Altman DG, Gotzsche PC, Juni P, Moher D, Oxman AD, et al. The Cochrane Collaboration's tool for assessing risk of bias in randomised trials. BMJ 2011;343:d5928. Higgins JPT, Green S. Cochrane handbook for systematic reviews of interventions. 1st ed. West Sussex, England: John Wiley & Sons Ltd; 2008. Homayoun H, Moghaddam B. NMDA receptor hypofunction produces opposite effects on prefrontal cortex interneurons and pyramidal neurons. J Neurosci 2007;27:11496e500. Jarventausta K, Chrapek W, Kampman O, Tuohimaa K, Bjorkqvist M, Hakkinen H, et al. Effects of S-ketamine as an anesthetic adjuvant to propofol on treatment response to electroconvulsive therapy in treatment-resistant depression: a randomized pilot study. J ECT 2013;29:158e61. Kamiya Y, Andoh T, Furuya R, Hattori S, Watanabe I, Sasaki T, et al. Comparison of the effects of convulsant and depressant barbiturate stereoisomers on AMPAtype glutamate receptors. Anesthesiology 1999;90:1704e13. Kellner CH, Briggs MC, Pasculli RM, Bryson EO. Antidepressant effect of the first electroconvulsive therapy with ketamine and/or propofol. J ECT 2013;29:149. Kellner CH, Knapp R, Husain MM, Rasmussen K, Sampson S, Cullum M, et al. Bifrontal, bitemporal and right unilateral electrode placement in ECT: randomised trial. Br J Psychiatry 2010;196:226e34. Koike H, Iijima M, Chaki S. Involvement of AMPA receptor in both the rapid and sustained antidepressant-like effects of ketamine in animal models of depression. Behav Brain Res 2011;224:107e11. Krystal AD, Dean MD, Weiner RD, Tramontozzi 3rd LA, Connor KM, Lindahl VH, et al. ECT stimulus intensity: are present ECT devices too limited? Am J Psychiatry 2000;157:963e7. Krystal AD, Weiner RD. ECT seizure therapeutic adequacy. Convuls Ther 1994;10: 153e64. Krystal AD, Weiner RD, Dean MD, Lindahl VH, Tramontozzi 3rd LA, Falcone G, et al. Comparison of seizure duration, ictal EEG, and cognitive effects of ketamine and methohexital anesthesia with ECT. J Neuropsychiatry Clin Neurosci 2003;15: 27e34. Kudoh A, Takahira Y, Katagai H, Takazawa T. Small-dose ketamine improves the postoperative state of depressed patients. Anesth Analg 2002;95:114e8. table of contents. Lai R, Katalinic N, Glue P, Somogyi AA, Mitchell PB, Leyden J, et al. Pilot doseresponse trial of i.v. ketamine in treatment-resistant depression. World J Biol Psychiatry 2014;15:579e84.

Lapidus KA, Levitch CF, Perez AM, Brallier JW, Parides MK, Soleimani L, et al. A randomized controlled trial of intranasal ketamine in major depressive disorder. Biol Psychiatry 2014;76:970e6. Loo C, Simpson B, MacPherson R. Augmentation strategies in electroconvulsive therapy. J ECT 2010;26:202e7. Loo CK, Katalinic N, Garfield JB, Sainsbury K, Hadzi-Pavlovic D, Mac-Pherson R. Neuropsychological and mood effects of ketamine in electroconvulsive therapy: a randomised controlled trial. J Affect Disord 2012;142:233e40. McDaniel WW, Sahota AK, Vyas BV, Laguerta N, Hategan L, Oswald J. Ketamine appears associated with better word recall than etomidate after a course of 6 electroconvulsive therapies. J ECT 2006;22:103e6. McGirr A, Berlim MT. A comment on Fond and colleagues' systematic review and meta-analysis of ketamine in the treatment of depressive disorders (Psychopharmacology 2014; Jul 20 [Epub ahead of print]). Psychopharmacol (Berl) 2014;231:3907e8. McGirr A, Berlim MT, Bond DJ, Fleck MP, Yatham LN, Lam RW. A systematic review and meta-analysis of randomized double-blind controlled trials of ketamine in the rapid treatment of major depressive episodes. Psychol Med 2015:1e12. 2014 Jul 10 [Epub ahead of print], http://dx.doi.org/10.1017/ S0033291714001603 [in press]. Montgomery SA, Asberg M. A new depression scale designed to be sensitive to change. Br J Psychiatry 1979;134:382e9. Murrough JW, Burdick KE, Levitch CF, Perez AM, Brallier JW, Chang LC, et al. Neurocognitive effects of ketamine and association with antidepressant response in individuals with treatment-resistant depression: a randomized controlled trial. Neuropsychopharmacology 2014. http://dx.doi.org/10.1038/npp.2014.298. epub: 2014 Nov 6. Murrough JW, Iosifescu DV, Chang LC, Al Jurdi RK, Green CE, Perez AM, et al. Antidepressant efficacy of ketamine in treatment-resistant major depression: a two-site randomized controlled trial. Am J Psychiatry 2013;170:1134e42. Okamoto N, Nakai T, Sakamoto K, Nagafusa Y, Higuchi T, Nishikawa T. Rapid antidepressant effect of ketamine anesthesia during electroconvulsive therapy of treatment-resistant depression: comparing ketamine and propofol anesthesia. J ECT 2010;26:223e7. Ostroff R, Gonzales M, Sanacora G. Antidepressant effect of ketamine during ECT. Am J Psychiatry 2005;162:1385e6. Petrides G, Fink M, Husain MM, Knapp RG, Rush AJ, Mueller M, et al. ECT remission rates in psychotic versus nonpsychotic depressed patients: a report from CORE. J ECT 2001;17:244e53. Prudic J, Haskett RF, Mulsant B, Malone KM, Pettinati HM, Stephens S, et al. Resistance to antidepressant medications and short-term clinical response to ECT. Am J Psychiatry 1996;153:985e92. Rasmussen KG, Kung S, Lapid MI, Oesterle TS, Geske JR, Nuttall GA, et al. A randomized comparison of ketamine versus methohexital anesthesia in electroconvulsive therapy. Psychiatry Res 2014;215:362e5. Rasmussen KG, Ritter MJ. Some considerations of the tolerability of ketamine for ECT anesthesia: a case series and review of the literature. J ECT 2014;30:283e6. Riley RD, Higgins JP, Deeks JJ. Interpretation of random effects meta-analyses. BMJ 2011;342:d549. Rosenthal R. Meta-analytic procedures for social research. Newbury Park, CA: Sage Publications; 1993. Sackeim HA, Dillingham EM, Prudic J, Cooper T, McCall WV, Rosenquist P, et al. Effect of concomitant pharmacotherapy on electroconvulsive therapy outcomes: short-term efficacy and adverse effects. Arch Gen Psychiatry 2009;66: 729e37. Sos P, Klirova M, Novak T, Kohutova B, Horacek J, Palenicek T. Relationship of ketamine's antidepressant and psychotomimetic effects in unipolar depression. Neuroendocrinol Lett 2013;34:287e93. Wang X. Antidepressant effect of the first electroconvulsive therapy with ketamine and/or propofol: response. J ECT 2013;29:149. Wang X, Chen Y, Zhou X, Liu F, Zhang T, Zhang C. Effects of propofol and ketamine as combined anesthesia for electroconvulsive therapy in patients with depressive disorder. J ECT 2012;28:128e32. Whitehouse AM, Scott AI. Monitoring seizure activity. In: Freeman CP, editor. The ECT handbook. London, UK: The Royal College of Psychiatrists; 2005. WHO. The ICD-10 classification of mental and behavioural disorders: clinical descriptions and diagnostic guidelines. 10th ed. Geneva, Switzerland: World Health Organization; 1992. Yalcin S, Aydogan H, Selek S, Kucuk A, Yuce HH, Karababa F, et al. Ketofol in electroconvulsive therapy anesthesia: two stones for one bird. J Anesth 2012;26: 562e7. Yoosefi A, Sepehri AS, Kargar M, Akhondzadeh S, Sadeghi M, Rafei A, et al. Comparing effects of ketamine and thiopental administration during electroconvulsive therapy in patients with major depressive disorder: a randomized, double-blind study. J ECT 2014;30:15e21. Zarate Jr CA, Brutsche NE, Ibrahim L, Franco-Chaves J, Diazgranados N, Cravchik A, et al. Replication of ketamine's antidepressant efficacy in bipolar depression: a randomized controlled add-on trial. Biol Psychiatry 2012;71:939e46. Zarate Jr CA, Singh JB, Carlson PJ, Brutsche NE, Ameli R, Luckenbaugh DA, et al. A randomized trial of an N-methyl-D-aspartate antagonist in treatmentresistant major depression. Arch Gen Psychiatry 2006;63:856e64.