#+REVEAL_ROOT: https://cdn.jsdelivr.net/npm/reveal.js
#+REVEAL_ROOT: ./reveal.js
#+REVEAL_TITLE_SLIDE:
%t
%a
%d
#+OPTIONS: toc:1 num:nil
#+AUTHOR: Armin Moradi
#+TITLE: Ethosuximide in MDD
#+DATE: 2020-02-11
* Overview
** Introduction
- ? A new rapid-acting antidepressant akin to ketamine
#+BEGIN_NOTES
Results of a previous preclinical study showed that ethosuximide which is an anticonvulsant drug can work as a rapid-acting antidepressant agent akin to ketamine. This study attempted to study this effect more closely. So far, effective rapid-acting antidepressants have been hard to come by, so the promise of a new agent was exciting enough that this group decided to study it further in a multicenter RCT.
#+END_NOTES
** Rationale (why care?)
- already widely used as anticonvulsant agent
- known and better side-effect profile vs ketamine and esketamine
- rapid antidepressant action in rodents
- potential for use in treatment-resistant MDD (both es/ketamine)
#+BEGIN_NOTES
ketamine has quite a few side effects, some of which are psychotomimetic effects (seen in shoulder dislocations! quite humorous), dissociation (it's used in ER), and there is potential for abuse (special K)
#+END_NOTES
** Rationale (but...how?)
- Lateral Habenula (LHb) burst fire inhibition = rapid-acting antidepressant
- NMDA + T-VSCC = burst fire
- A AND B = C -> inhibit A OR inhibit B = inhibit C
#+BEGIN_NOTES
- Habenula previously referred to the stalk of the pineal gland, but now refers to a group of cells around the gland
- Lateral Habenula (LHb) burst fire inhibition has been associated with rapid-acting antidepressant effect of es/ketamine
- LHb requires both N-methyl-D-aspartate (NMDA) receptor antagonist (ketamine) and T-type low-voltage-sensitive calcium channel (T-VSCC) (ethosuximide)
- Therefore, ethosuximide (T-VSCC inhibitor) theoretically affects the same pathway. The thinking goes if 2 components are necessary to initiate LHb burst fire, inhibiting either T-VSCC or NMDAR should inhibit the pathway, meaning es/ketamine and ethosuximide should theoretically be equivalent. Spoiler: they're not.
#+END_NOTES
** Primary Goals
- Is ethosuximide any good for depression?
- Spoiler: no!
#+BEGIN_NOTES
The primary outcome was Hamilton depression rating scale score change in 5 hours and 2 weeks interval post administration. the 2 week interval was after repeated administration.
There were secondary outcomes too: changes in Montgomery depression scale (MADRS)
Hamilton anxiety scale was also used and monitored for change, although they did not specify it as an outcome
In this case, we're talking about change in score vs placebo
#+END_NOTES
** Methods
- Double-blind
- Randomized
- Placebo controlled
- Three mental health centres in China
- Three different doses
- 80 participants, 4 groups (3 ethosuximide, 1 placebo)
#+BEGIN_NOTES
more details later
#+END_NOTES
** Results
- No dice!
#+BEGIN_NOTES
no significant change in any of the scores monitored
#+END_NOTES
** Author's Conclusion
- No good!
#+BEGIN_NOTES
The results speak for themselves. The authors concluded that ethosuximide does not produce the robust antidepressant effect that we expect from es/ketamine.
Theoretically sound, but practically not so.
So let's dive deeper into the study itself.
#+END_NOTES
* Diving head first!
* Authors
** Funding and Support
- National Nature Science Foundation of China
- Wuxi Technology Bureau
- Wuxi Municipal Health and Family Planning commission
** Conflict of interest
- One researcher (Dr. K. Hashimoto) has a pending patent for "Application of R-ketamine and salt thereof as pharmaceuticals" by Chiba Univesity, and has received research support by Otsuka, Sumitomo-Dainippon, and Taisho pharmaceutical companies.
* Introduction
** Background and rationale
- Reference to prior preclinical studies.
- Information regarding the role and mechanism of action of ketamine and its relation to ethosuximide was provided
- Motivation: good parts of ketamine less the bad parts
#+BEGIN_NOTES
- Initial idea coming from a preclinical study on rodents was briefly mentioned, although it was a bit difficult to find that paper
- Brief history of ketamine and its off-label use and final USFDA approval in Mar 2019 were discussed
- Ketamine and ethosuximide in terms of their clinically-relevant potential similarities and differences were discussed
- More foundational topics like mechanism of action of each compound were discussed
- Motivation behind the study was also elaborated by discussing desirable properties of ketamine (rapid-acting, treatment-resistant MDD) alongside its drawbacks (side-effects)
- Overall, the study was well justified and had a clear goal
#+END_NOTES
** Hypothesis and research question
- Hypothesis: ethosuximide is efficacious for treatment of MDD in treatment-naive patients
- Objective: to evaluate the antidepressant efficacy of ethosuximide in comparison with placebo in treatment-naive adult patients with MDD
- Primary outcome: HAM-D score change from baseline vs placebo over 5h and 2 weeks
- Secondary outcome: MADRS score change from baseline vs placebo over 5h and 2 weeks
#+BEGIN_NOTES
Hypothesis was succinct and the study was fairly close to the hypothesis mentioned.
The introduction seemed to emphasize unique properties of ketamine and potentially ethosuximide to generate rapid antidepressant effects and their use in treatment-resistant MDD, but the hypothesis, objective, and primary and secondary outcomes didn't seem to reflect that emphasis. The study was essentially focused on testing whether ethosuximide was worth further research.
Generally, the purpose and design were clear and well-followed.
#+END_NOTES
* Methods
** Design
- Multicenter, Double-blind, Placebo-controlled
- First RCT for ethosuximide: no sample size estimation
#+BEGIN_NOTES
- Wuxi Mental Health Center, Anhui Mental Health Center, Chaohu Hospital of Anhui Medical University
- Effect size vs intervention was unclear as it's the first RCT re ethosuximide in MDD, so sample size could not be estimated based on statistics. So, based on animal studies, they estimated that a sample size of 20 would be adequate for this initial study. The authors acknowledged this to be an exploratory study.
#+END_NOTES
** Recruitment
- Adults 18-65 with confirmed single or recurrent MDD based on MINI
- No information on method of recruitment was provided
- No information on compensation
#+BEGIN_NOTES
MINI: MiniInternational Neuropsychiatric Interview
The study included stable, treatment-"naive" adults and excluded pretty much everyone else.
Not sure what "serologic findings" means in this case.
Also, no explanation as to why these groups were excluded was given. I suspect that with their previous hint that this was an exploratory study, they wanted to reduce the analysis surface by excluding as much variation as possible, but that's just a guess. Would have been nice if they included some brief explanation.
#+END_NOTES
** Screening
- No pre-screening information was provided
- inclusion criteria
- score of 18 on 17-item HAM-D _AND_ no antidepressant treatments in past 6 months
- Exclusion criteria
- substance abuse/treatment in previous 3 months
- active suicidal intent
- pregnancy
- bipolar
- psychotic disorder or current psychotic symptoms
- unstable medical illness
- significant neurological illness
- abnormal serologic findings
** Participants
- 112 patients assessed, 17 declined to participate, 32 did not meet inclusion criteria
- 80 randomly assigned evenly to 4 groups known only to pharmacy administrators (double-blind)
- Ethosuximide 500mg
- Ethosuximide 1000mg
- Ethosuximide 1500mg
- Placebo
- even distribution
** Administration
- Each group has x1 dose on first day of treatment
- 1500mg and placebo groups had repeat dose daily for 2 weeks
- Timing of doses in the day, preparation of oral medication (pill, tablet, liquid, gel, etc.) not discussed
#+BEGIN_NOTES
the wording and explanation of the administration regimen could definitely improve. They referred to having 4 groups with single dose on day 1, then 2 groups with single dose at 2 week mark, then daily dose x2 weeks for 2 groups, yet their results only include 4 groups of 20 participants
#+END_NOTES
** Assessment
- HAM-D and MADRS were used for scoring
- Visual Analog Scale (VAS) was used to monitor dependent symptoms
- HAM-D, HAM-A, MADRS, and VAS were administered at 1h and 5h after single dose oral administration
- for repeat group, HAM-D, MADRS, HAM-A, and VAS were measured at
- baseline
- end of first week
- end of second week
#+BEGIN_NOTES
Previous study of male adults showed peak plasma concentration levels of 15ug/mL 3-5hr after single 750mg administration, so 1h and 5h were chosen for this study to capture peak plasma concentration in the scoring scales
#+END_NOTES
** Measures
- HAM-D, HAM-A, MADRS, VAS
- General demographics data:
- age
- gender
- years of education
- height
- weight
- length of MDE
** Statistical Analysis
- data collected as means +/- std deviation
- Baseline characteristics compared using Pearson x^2 test for categorical variables
- One-way analysis of variance (ANOVA) used for continuous variables
- primary outcome: HAM-D
- secondary outcome: MADRS, HAM-A, and VAS
- changes in baseline in above measures were assessed with repeated measures one-way ANOVAs followed by post-hoc Fisher's least significant difference test
-
* Results
** Scores
| | Control | 500mg | 1000mg | 1500mg |
|-------+---------------+---------------+---------------+---------------|
| HAM-D | 25.20 +- 5.75 | 26.15 +- 6.56 | 23.95 +- 4.37 | 26.85 +- 4.06 |
| HAM-A | 10.35 +- 2.18 | 11.30 +- 2.60 | 10.95 +- 2.63 | 10.65 +- 2.01 |
| MADRS | 26.90 +- 5.71 | 28.20 +- 6.84 | 26.00 +- 4.65 | 28.90 +- 4.19 |
** Statistical Significant
- Primary outcome (HAM-D)
- 5h -> P=0.243
- 2w -> P=0.263
- Secondary outcome
- MADRS
- 5h -> P=0.214
- 2w -> P=0.260
- HAM-A
- 5h -> P=0.425
- 2w -> P=0.824
- VAS
- 5h -> P=0.472
- 2w -> P=0.623
- No significance
#+BEGIN_NOTES
The proposed P-value for statistical significance was 0.05, but all of these P-values are at least 4 times higher.
#+END_NOTES
** Clinical Significance
?
#+BEGIN_NOTES
At this point it's unlikely that ethosuximide is something we should keep our eye on. More studies are needed (with one under way in China).
For now, though, no change in clinical practice is warranted.
#+END_NOTES
** Reliability
- The good
- Primary outcome based on objective measures (HAM-D, HAM-A, MADRS, VAS)
- 100% rate of completion
- Double blind
- Similar groups at the start of the study, treated equally except treatment
- All patients analyzed
- Low (<40%) placebo response
- The bad
- Sample size assumed based on animal studies, no statistical analysis
- The ugly
- ...
Generally Reliable
#+BEGIN_NOTES
There are many things that are done well about this study
#+END_NOTES
* Conclusion
** Statistical Interpretation
#+BEGIN_QUOTE
Results indicate that ethosuximide does not have antidepressant actions in non-medicated patients with MDD
#+END_QUOTE
- There isn't much room for artistic interpretation
- Both primary and secondary outcomes: large P-value = No significance
- High placebo response
- >40% response muddies the water
- this study -> low placebo response
** Limitations
- Only 80 patients -> "preliminary results"
- Needs bigger samples
- Strict inclusion criteria
- "unclear whether findings will generalize to all patients with treatment-resistant MDD."
** Conclusion
- No severe adverse events
- Did not mention what adverse events, if any, were reported by patients
- Did not alter VAS score after 2 weeks -> low potential for abuse
** Conclusion (2)
- Demonstrated antidepressant effect of ethosuximide in rats with absence epilepsy and comorbid depression-like behaviour
- Also in rats subjected to chronic restraint stress
#+BEGIN_QUOTE
It seems that antidepressant-like effects of ethosuximide are limited to WAG/Rij rats, which suggests a link between absence seizures and depressive-like behaviours in this strain of rats.
#+END_QUOTE
** Conclusion (3)
- Previous studies failed to produce antidepressant effect in a stress model of chronic social defeat
- R-ketamine did show antidepressant effect
- This study supports the above study
#+BEGIN_NOTES
It is known that, according to a meta-analysis, >40% placebo response can undermine detection of statistical significat difference between active drug and placebo. The authors considered this effect and report a low placebo response which corroborates with the results.
#+END_NOTES
** Conclusion (4)
#+BEGIN_QUOTE
We could not find any antidepressant effect of ethosuximide on the clinician-reported HAM-D and MADRS scores after single and repeated oral administration. Therefore, it is unlikely that ethosuximide elicits ketamine-like rapid-acting antidepressant actions in patients with MDD, but the negative findings need to be replicated by other research groups.
#+END_QUOTE
* Discussion
- Results appropriately interpreted
- Adequately explained limitations except
- Standardization process among 3 centers
- Medication administration process
- Consistent conclusion
- Generalizable? No
- Ethnicity was not discussed
- Use of ethosuximide in treatment-naive patients is unlikely at this point
- Literature correspondence? Yes
- R-ketamine vs ethosuximide in rats
- doi:10.1016/j.pnpbp.2019.109652
- doi:10.1093/ijnp/pyy065
- doi:10.1093/ijnp/pyy072
* Clinical practice
- No change at this point
* Future
- Ethosuximide in treatment-resistant MDD
- Bigger sample size
* QA?