Package Insert For Epidiolex - Does It Suggest A Problem With Medical Cannabis?

Cannabidiol (C₂₁H₃₀O₂)

Epidiolex was approved by the FDA two weeks ago for Lennox-Gastaut syndrome or Dravet syndrome in patients 2 years of age and older.  Epidiolex is cannabidiol (abbreviated CBD) one of several compounds in the plant Cannabis sativa.   I had a previous post on this compound but that was before the package insert came out.  I like to study package inserts of all of the pharmaceuticals I encounter to prevent unexpected side effects, anticipate drug interactions, look at the current prescribing recommendations, and study all of the safety considerations.  Every drug has a section in that package insert about the pharmacokinetics, pharmacodynamics, and considerations in the case of hepatic or renal impairment.  In some cases there are very specific recommendations for dosing with metabolic impairment or potential drug-drug interactions.  The other interesting aspect in this case is that Epidiolex is considered the first botanical extract to be FDA approved and the first cannabis derived compound.  A significant part of the population considers cannabis to be a benign natural product with none of the usual pharmaceutical concern about organ toxicity and drug interactions.

Reading the actual package insert a few things jump out at me today.  The original indications are the same, but the logical question is whether this medication will be used for off labeling prescribing for other indications.  After hearing one of the top epileptologists  in state talk about the use of cannabinoids for epilepsy, there is also the question of whether the diagnosis is correct.  In that lecture he pointed out that a case example in the news media probably did not have the diagnosis and that the expert in the state who could make that diagnosis was not consulted.

The dosing of the drug is fairly robust going from 5 mg/kg/day up to 20 mg/kg/day.  For a 70 kg man that comes out to a max dose of 1400 mg/day putting it in the range of several other anticonvulsants from different classes.

There are warnings about hepatotoxicity.  Early in the document, it states that some patients will experience elevated liver function tests and in some cases with develop overt side effects leading to drug discontinuation.  Baseline screening is recommended with AST, ALT, and total bilirubin.  Patients with elevated baseline transaminases were more likely to experience further elevation of these tests than those subjects with no baseline elevation.  The Child Pugh classification of severity of liver disease is used as a metric with dose adjustments suggested for mild, moderate, and severe disease.

Thirteen percent of patients had ALT elevations that were three times the upper limit of normal (ULN).  Less than 1% had transaminases that were 20 times the ULN and some patients were hospitalized.  In a third of the cases the transaminase elevation resolved without treatment.  In the other two thirds it resolved with discontinuation of the Epidiolex or the associated anticonvulsant (valproate).

Risk factors (associated drugs - clobazam, valproate), dose, and baseline transaminases) were discussed as well as monitoring.  Given the prevalence of the problem screening transaminases at 1 month, 3 months, and 6 months and as indicated after that.  More importantly - screening for the physical illness from drug induced liver disease ("explained nausea, vomiting, right upper quadrant abdominal pain, fatigue, anorexia, or jaundice and/or dark urine") can lead to further evaluation.  Three scenarios for discontinuing the Epidiolex are recommended:

1.  Transaminase levels greater than 3 times the ULN.

2.  Bilirubin levels greater than 2 times the ULN.

3.  Transaminase levels greater than 5 times the ULN. 

My read of the difference between 1 and 3 is that 1 can be a temporary measure but 3 should be permanent.  In my experience with valproate, I would definitely discontinue with these levels.  That is based on the well validated concern that valproate can cause significant hepatotoxicity. It is still possible that additional trials and post marketing surveillance will show that there is not long term concern with CBD.  In the trials transaminase elevation was the most frequent reason that the drug was discontinued (24% versus 3% on placebo).

Drug interactions noted that could be clinically significant. Epidiolex is metabolized by  CYP3A4 and CYP2C19 so that inhibitors of these enzymes can potentially increase the plasma levels.  Strong inhibitors of CYP3A4 include HIV antivirals, antifungals, and buprenorphine.  There are no strong CYP2C19 inhibitors.

Inducers of the same enzymes can lower Epidiolex levels and the standard inducers of those enzymes are carbamazepine, oxcarbazapine, phenytoin, HIV antivirals, prednisone and glucocorticoids, and St. John's Wort. 

Additional warnings about the use of the drug include somnolence and sedation (32%), suicidal ideation and behavior, hypersensitivity reactions, and the risk of withdrawing an anticonvulsant and need to do it gradually.   Regarding the suicidal ideation and behavior the only data presented was from a large (N=199) pooled analysis of clinical trials.  It is a standard warning on all anticonvulsant drugs and there was nothing specific to Epidiolex or CBD.

Clearly Epidiolex or CBD extracted and used at pharmaceutical doses may have some of the power of pharmaceuticals but also has the same significant side effects.  The side effect profile and drug interaction concerns are very similar to other pharmaceuticals that are used to treat epilepsy. This raises some interesting issues in states like Minnesota where high potency extracts of cannabis are being sold as medical cannabis and there is minimal medical supervision - primarily because there is scant evidence that cannabis extracts are medical treatments.  As I previously observed from the most recent report of the Minnesota medical cannabis program, extracts are being sold in this state that result in the ingestion of 12.2 - 1,439.2 mg/day of CBD.  The middle to high end of that range is clearly in the dose range for Epidiolex and the extracts are not prescribed or monitored by physicians - at least there is no requirement for that to happen.  Looking at all of the available data it is clear that the person taking 1,439.2 mg/day is an outlier and the next cluster of patients is at the 100-200 mg/day range. 

In Minnesota, a medical provider certifies a patient as having a condition that qualifies them for medical cannabis. In the case of this report it is chronic pain.  That patient goes to a medical cannabis dispensary and discusses what they want with a pharmacist.  In the case of high CBD products, as far as I know there is no recommended screening, monitoring, or patient education.  Just based on what I read in the current Epidiolex package insert, if the CBD content of the medical cannabis is in a similar dose range that is the equivalent of taking a new pharmaceutical and making it an over the counter drug.  The neurologists prescribing Epidiolex have good guidance on what needs to be monitored and are undoubtedly very familiar with the compound.  Other physicians including psychiatrists need an awareness of the pharmacology of CBD - especially if the dose is in the range suggested by this package insert.

If it was needed, this seems like further evidence that the miracle of medical cannabis has affected the judgment of many who seem to consider it a benign natural product. It turns out in this case, it can have a therapeutic effect on specific seizures at a significant dose for conditions that did not have many good options.  That treatment comes with clear risks.  The risk is reduced since all of the patients treated for the indicated seizure disorders are being followed by neurologists who specialize in the polypharmacy necessary to treat complex seizure disorders.  That includes monitoring potential drug interactions and toxic effects.  Can we say the same thing for people obtaining it through the medical cannabis program or being prescribed the drug off label?

Medical cannabis needs to be taken as seriously for the side effects as it does for the purported benefits.


George Dawson, MD, DFAPA


Reference:

Full Prescribing Information for Epidiolex. FDA approved package insert.


Supplementary 1:

Any FDA package insert is available online by Googling:  "[Drug name] FDA Package Insert"   The PDF of that drug insert will pop up and you will have access to same the full prescribing information that any physician has.

Supplementary 2:

In Minnesota, there are two companies that are the exclusive providers of non-smokable medical cannabis products Leafline Labs and Minnesota Medical Solutions.  Actual THC and CBD content is available on the web sites of both companies.

Leafline Labs has a vaporization product, a sublingual spray, an oral suspension, and a topical preparation.  The highest concentration of CBD in the oral solution is 20 mg/ml.  Epidiolex is 100 mg/ml.

Minnesota Medical Solutions has similar delivery forms and their oral products are capsules and solutions in both 47.5 mg CBD or 100 mg per milliliter CBD.  The latter is the same as prescription strength Epidiolex.

First FDA Approved Cannabis Product - Epidiolex

Cannabidiol (CBD)

Headlines yesterday were that the FDA Peripheral and Central Nervous System Drugs Advisory Committee approved a cannabis product for the treatment of seizures.  The draft agenda for that meeting is available on the FDA web site but no vote or minutes of the meeting.  This has been a widely hyped application of medical cannabis since the term was first used to try to start the legalization of cannabis products.  Ironically the first example of a patient used as an example for this application did not have the syndrome in question.

At the  time I am typing this post there is no FDA approved package insert for the compound and no prescribing information available on the manufacturer's website.  There is a considerable amount of detail in the patent that is available online.  The patent focuses on the use of cannabidiol in treatment resistant epilepsy including Dravet syndrome; myoclonic absence seizures or febrile infection related epilepsy syndrome (FIRES). The patient states that when  cannabidiol is used for these disorders the total reduction in seizure frequency is 50-70% but as high as 90-100%.  The product is formulated to be used separately or with other anti-epileptic drugs (AED).  The patent suggests that the formulation is comprised of highly purified cannabidiol extract (98% w/w) with most of the psychoactive components - tetrahydrocannabinol (THC) removed.

The epidemiology and treatment of epilepsy especially Dravet's syndrome is reviewed in the patent.  It is clear from that data that there are no good treatments for these disorders and that conventional AEDs are partially effective at best.  The four treatment scenarios (non of which are randomized controlled clinical trials) all demonstrate efficacy for CBD in intractable epilepsy.  These results are included in the table below.

Trials	Treatment	General Response	Best Response
N=27 children and young adults Dx: severe childhood onset treatment resistant epilepsy (TRE)	Initial dose: CBD 5mg/kg/day titrated to a max of 25 mg/kg/day in addition to baseline AEDs	-after 3 months of therapy, 48% of patients had an equal to or greater than >50% reduction in seizures. -	- 2 patients were seizure free at three months - 6 patients had a 90% reduction in seizures at three months
N=9 children and young adults with treatment-resistant Dravet syndrome	- as above	- after 3 months of therapy, 56% of patients had an equal to or greater than 50% reduction in seizures,	- 2 patient were seizure free at three months -3 patients has a 90% reduction of seizures at three months
N=4 patients with myoclonic absence seizures who were taking at least two concomitant anti-epileptic drugs	-as above	after 3 months of therapy, 2 of the patients had an equal to or greater than 50% reduction in seizures, 1 patient (25%) had a 90% reduction at the three month stage.	-0 patients were  seizure free -1 patients with a 90% reduction of seizures at three months
N=3 patients with FIRES (febrile infection related epilepsy syndrome)	-as above	- 2/3 children has a 90% reduction in seizures	-2/3 children had a 90% reduction in seizures and regained some motor and intellectual functioning.

In these studies no patients were withdrawn because of side effects.  Common adverse events were somnolence, fatigue, decreased appetite, increased appetite and diarrhea.  Given the degree of polypharmacy it is probably difficult to determine what side effects are due to CBD and either the other compounds or interactions with the other compounds.   All patients had severe epilepsy with hundreds to thousands of seizures per month and in many cases required episodic aggressive treatment to stop status epilepticus.  In this population, placebo response would be expected to be very low.  The use of CBD in this population could easily be justified on a compassionate use basis because of illness severity and the lack of any severe complications from the CBD.  The results of these open label trials appear to be complied in reference 1.  Some of the authors have also published a review and  randomized, placebo controlled dose ranging study of the safety of CBD.  I cannot imagine that there was not an efficacy arm to that study.  There is also a 2017 commentary that looks at the general question about whether there is adequate evidence that CBD is effective for epilepsy(4).

A mysterious part of the patent process is that this is a simple organic chemistry extraction of CBD from cannabis plants and then resuspending for oral dosing.  The authors specify that THC is essentially removed.  The dosing is done in a standard mg/kg dose that would be expected for any medication. That dose range appears to be part of the patient. Does this patent mean that any CBD extracted from cannabis is the intellectual property of this company or is there more to it than that?  If that is the case then any product with a CBD/THC ratio could be patented and there are currently 11 of these compounds prepared by a similar extraction process in the Minnesota Medical Cannabis Program.  Is each one of these compounds patentable? Or is there some additional information about the formulation that is not included in the patent that would make this formulation more unique? At one point the patent states that the CBD can be synthesized as well as extracted.

The market for this product is potentially significant, if it works in less severe forms of epilepsy. Will it work for example in the case of a person who has a few seizures per month after their AEDs have been optimized?  Can that person take the necessary dose of CBD and not experience significant side effects?  Will the company or other companies try to get another FDA approved indication for CBD related products from the FDA? I suspect the standard will be higher in patients where there are compounds of equal or better efficacy.  Will physicians prescribe CBD derived drugs off-label to people requesting them for philosophical reasons (eg. I want to be treated with a substance derived from a botanical)?  Will the positive effects noted in these small trials persist as more and more people with epilepsy are exposed to the drug?  Will there be more significant side effects with post marketing surveillance and maintenance use? There are still a significant number of people who do not appear to respond very well to CBD.  Those are some of the questions I had considering the implications of this product release.

On theoretical grounds, a key question is whether the utility of this compound in severe epilepsy will lead to additional drug discovery of more compounds for this difficult to treat problem.  For now, the FDA has made a sound decision getting this compound into the hands of the physicians who treat these disorders on a day to day basis.


George Dawson, MD, DFAPA


References:


1:  Devinsky O, Marsh E, Friedman D, Thiele E, Laux L, Sullivan J, Miller I, Flamini R, Wilfong A, Filloux F, Wong M, Tilton N, Bruno P, Bluvstein J, Hedlund J, Kamens R, Maclean J, Nangia S, Singhal NS, Wilson CA, Patel A, Cilio MR. Cannabidiol in patients with treatment-resistant epilepsy: an open-label interventional trial. Lancet Neurol. 2016 Mar;15(3):270-8. doi: 10.1016/S1474-4422(15)00379-8. Epub 2015 Dec 24. Erratum in: Lancet Neurol. 2016 Apr;15(4):352. PubMed PMID: 26724101.

2:  O'Connell BK, Gloss D, Devinsky O. Cannabinoids in treatment-resistant epilepsy: A review. Epilepsy Behav. 2017 May;70(Pt B):341-348. doi: 10.1016/j.yebeh.2016.11.012. Epub 2017 Feb 8. Review. PubMed PMID: 28188044.

3: Devinsky O, Patel AD, Thiele EA, Wong MH, Appleton R, Harden CL, Greenwood S, Morrison G, Sommerville K; GWPCARE1 Part A Study Group. Randomized, dose-ranging safety trial of cannabidiol in Dravet syndrome. Neurology. 2018 Apr 3;90(14):e1204-e1211. doi: 10.1212/WNL.0000000000005254. Epub 2018 Mar 14. PubMedPMID: 29540584

4:  Perucca E. Cannabinoids in the Treatment of Epilepsy: Hard Evidence at Last? J Epilepsy Res. 2017 Dec 31;7(2):61-76. doi: 10.14581/jer.17012. eCollection 2017 Dec. Review. PubMed PMID: 29344464.

5:  FDA Briefing Document Peripheral and Central Nervous System Drugs Advisory Committee Meeting April 19, 2018 NDA 210365Cannabidiol

A Report From The Minnesota Medical Cannabis Program

I have described the Minnesota Medical Cannabis Program in a previous post.  It is a unique program because it does not provide smokable cannabis for medical purposes.  All of the cannabis is in a form that is vaporized, edible, absorbed through the oral mucosa, or transcutaneously after application to the skin.  It is produced by two companies and made into products of varying amounts of  tetrahydrocannabinol (THC), cannabidiol (CBD) .  In the original matrix from the first post there appeared to be a total of about 11 products based on the THC and CBD ratios. For the purpose of this report the cannabis products were classified base on ratios of cannabinoids and route of administration according to the following table:

The program came out with a report on the experience of patients with chronic intractable pain using the program this week.  They produce frequent reports and this is the latest. Although it is not a scientific study of the associated issues of pain relief it is interesting because of a number of considerations not the least of which is what happens when a state starts to manage a CSA Schedule 1 drug on their own and come up with their own process for certification and use.

The report details the cohort of patients who were qualified on the basis of intractable pain according to the following definition:  “pain whose cause cannot be removed and, according to generally accepted medical practice, the full range of pain management modalities appropriate for this patient has been used without adequate result or with intolerable side effects.” (p 4).

The report focused on the results and side effects of 2245 patients certified by various practitioners for intractable pain between August 1, 2016 and December 31, 2016. It contains basic demographics of the patients and practitioners as well as a breakdown of the possible origins of their chronic pain.  All of these details are available in the original report and I encourage any interested reader to find the details there.  I am interested in the type of medical cannabis being used and the outcomes.

These details are fairly interesting from a descriptive standpoint. For example, just using the product definitions there is a distribution of how the products are purchased.  The following bar graph illustrates the distribution of purchased products by THC and CBD content:

 
The report gets into more specific details about the types of THC and CBD products in each category and the average ingestion of these products per day in milligrams (mg). I am reproducing the first of a 4 page table here that contains progressively fewer patients in each strata:

It is apparent that very high THC products that are inhaled or ingested predominate the product distribution.  Although the largest single group of patients were averaging 81.5 mg/day THC and 0.6 mg/day CBD the range is significant from 4.4 to 553.8 mg/day THC and 12.2 to 1,439.2 mg/day CBD.

Patient and treatment providers were asked to rate their degree of benefit from the medical cannabis program on a Likert scale where 1 = no benefit and 7 - great deal of benefit.  Using that system 61% of the patients rated their experience as a 6 or 7 compared with 43% of the healthcare professionals rating the patient benefit as a 6 or 7.  Specific benefits were rated and the top three included pain relief (56%), sleep improvement (10%), and reduction in pain medications and side effects (7%). 

An area of interest in the report is whether or not the patient is reducing the amounts of other pain medications used in response to the use of medical cannabis. This question is asked to the certifying health care professionals.  For this report, 586 responses of a total of 692 were used to determine that pain medications were reduced in 58% of the cases and not reduced in 48% of the cases.  221 reduced an opioid with 127/221 reducing the opioid by 50% or more.  In addition, 16 reduced a benzodiazepine and 128 reduced a pain medication "other than an opioid or benzodiazepine" - but benzodiazepines are not pain medications.  The full list of medications reduced or discontinued are available in Appendix E.  Minnesota does have a Prescription Monitoring Program for controlled substances making it possible to quantify and confirm all of this data rather than depending on survey results.

The section of the report that I found most interesting was the section on a standard group of 8 symptoms followed in this patient group for improvement or no improvement.  These symptoms are listed in the tables below.

The response options are on a standard analogue scale where 10 is the worst possible symptom and 0 is the symptom is not present.  The report listed the results of this scale applied to intractable pain patients as shown in the table below.

It is an interesting report in that it gives improvements in symptoms over 4 months as well as the percentage of patients a 30% improvement in symptoms at some point in the initial 4 months, the percentage of patients who had a 30% symptoms improvement in the 4 month follow-up period, and the percentage of patients who had the 30% symptoms improvement and maintained it for at least 4 months.  Only 11% of the pain patients maintained a 30% improvement in pain symptoms over the 4 month follow up despite higher ratings of initial pain relief.  The most significant improvements were in nausea and vomiting.

Side effects were reported in the final section with about 10% of patient reporting severe side effects.  These side effects included somnolence, sedation, headaches, dizziness, lightheadedness, confusion, fatigue, abdominal pain, mental fogginess, inability to concentrate, anxiety, panic attacks, and insomnia.  Physical side effects were reported roughly twice as often as mental side effects and most were in the mild to moderate range.  A typical metric found in clinical trials - medication discontinuation because of side effects was apparently not determined.  In a series of statements included in the report, one patient stopped because of a lack of efficacy and there was a suggestion that stopping could occur for that reason or financial reasons but the data was not clear.

Although the current report is focused on intractable pain a couple of things seems clear.  First, these reports are not scientific.  There are no comparison drugs or placebos.  Medical cannabis is added in many cases to a combination of existing opioid pain medications and benzodiazepines.  In a purely qualitative sense, they do show what products are preferred by customers and these products contain significant amounts of THC.  Second, in the case of the 8 symptom rating pain relief from medical cannabis appears to be modest at best.  A significant number of patients acknowledged severe side effects, but were these side effects of the same level of severity that might be seen in a clinical trial?  In those patients who replaced opioids or to a lesser extent benzodiazepines - was that because the initial medications were ineffective for pain or because cannabis has superior effects on pain?

The most interesting part of the data to me is that fact that medical cannabis is highly promoted for chronic pain.  That promotion was initially political - for the purpose of legalizing medical cannabis.  Currently it takes the form of cannabis saving people from opioid overdoses and this report makes an attempt to record reduced amounts of opioids use due to the cannabis. The problem is that it is all survey data.  There are no standardized doses of cannabis and no attempt to determine a placebo effect.  Physicians used to reading clinical trials need to ask themselves: Is a patient interested in using medical cannabis in Minnesota capable of answering the questions in an unbiased manner?  Will their interest/belief in medical cannabis influence survey results? And if that is true - what does it mean that medical cannabis ends up with such a poor result for pain relief over a period of 4 months?  I have concerns about the survey results based on my interviews with thousands of cannabis smokers.  Although I am seeing people with significant addiction problems, I don't see the side effect frequency of insomnia, anxiety, panic, and paranoia that I am used to hearing hearing about.  And in this group, I can't help but wonder how many of them have significant addictions? I also don't see a discussion of the fact that many opioid users commonly switch to cannabis when the opioid supply runs low or they make an attempt to stop using. There are potentially several mechanisms occurring in this population in addition to pain relief and side effects.

Another issue indirectly addressed by this report is what happens when you do an end run around the FDA?  I have certainly been a critic of the FDA and its regulatory processes, but in the end there is always a study available for public discussion.  That study typically has much more information content about drug efficacy and tolerability than the current Minnesota study because of the scientific design.  This report is the type of data that you get when that regulatory hurdle is ignored.

The direction and legacy of medical cannabis in Minnesota seems to be contingent on the status of recreational cannabis.  The program has been criticized for being too expensive compared with smokable cannabis.  If Minnesota legalizes recreational cannabis, that may be the preferred route by many of the people in this program. Questions about the efficacy of cannabis in intractable pain remain unanswered despite all of the details in this report.



George Dawson, MD, DFAPA


References:

1:  Minnesota Department of Health Office of Medical Cannabis.  Intractable Pain Patients in the Minnesota Medical Cannabis Program: Experience of Enrollees During the First Five Months. Report

2: Desrosiers NA, Ramaekers JG, Chauchard E, Gorelick DA, Huestis MA. Smoked cannabis' psychomotor and neurocognitive effects in occasional and frequent smokers. J Anal Toxicol. 2015 May;39(4):251-61. doi: 10.1093/jat/bkv012. Epub 2015 Mar 4. PubMed PMID: 25745105.


Graphics:

All tables excerpted from reference 1 as a public document with no copyright.