Slow COMT

Posted on September 25, 2026 by dulcetware

Catecholomethyltransferase (COMT) has long been known in neurobiology as an enzyme that breaks down dopamine in the synapse to prevent excessive post-synaptic stimulation and in neurology where inhibitors of the enzyme act to delay the peripheral breakdown of L-DOPA, so extending the effect of antiparkinsonian medications for an hour or two.

COMT methylates dopamine and thereby inactivates it. There has been found to be a polymorphism of COMT activity in humans without disease and this polymorphism can have cognitive effects. Older studies from the early 2000’s highlighted potential deficits of too high COMT activity in humans with the VAL158 (a valine amino acid in position 158) polymorphism because it makes the enzyme more thermostable. These effects include increased risk of psychiatric disorders such as schizophrenia and impairments on certain cognitive tasks. Broadly speaking, one could infer that a low dopaminergic state could impair reward associated performance and behaviour. Since COMT may be more critical in the prefrontal cortex than in the striatum, where there is more dopamine reuptake by dopamine transporter (DAT) as an alterative mechanism, it would not be surprising that any deficits would be cognitive and not motor (i.e. Parksonian-like).

However, we already see a wrinkle regarding schizophrenia, where traditional treatment has been with anti-dopaminergic medication, presumably worsening already low dopaminergic levels.

The alterative polymorphisms are heterozygote valine/methionine at this allele or homozygous methionine. Around 25% of Europeans are Met/Met (slow), 50% heterozygote, and 25% Val/Val (fast). In east Asian populations, Met tends to be lower.

Now there is much talk of slow COMT personality types and fast COMT types, roughly corresponding to type A and type B personalities. We have entered the realm of populist quasi-medicine.

Inevitably, there is a whole industry of what slow COMT types (because they are the ones who will obsessively read the literature) “need” to boost their COMT, ranging from avoiding “oestrogen mimickers” to avoiding coffee, tea, chocolate and red wine that “consume” COMT “resources”.

And there is a matching industry of genetic testing for finding your type and thereby, one might cynically say, achieving introspection. The genetic test costs around £150-200 to be done privately.

Finally, there is an industry of supplements, such as tyrosine (an amino acid simply present in protein food), key B vitamins (excess of B6 can give side effects), biopterin for fast COMT, and magnesium for slow COMT (a cofactor for COMT that is abundantly present in any living plant or animal cell that one might eat) because you tend to “burn through magnesium faster under stress”, presumably undergoing nuclear fusion into some other element instead of remaining as the major +2 cation in the intracellular compartment. There are also oestrogen processors that presumably reduce oestrogen levels and “stress nutrients”.

But there are a huge number of questions:

  • To what extent is COMT rate limiting? There is a long pathway of dopamine and NA breakdown. Does a 3 fold reduction in activity mean anything when the rate limiting step might be elsewhere?
  • COMT is not the only enzyme. What about monoamine oxidase (MAO) which performs the same function? Could there be compensatory mechanisms?
  • There might be other genetic polymorphisms that interact with COMT activity or indeed MAO activity.
  • Is there genetic linkage (a neighbouring gene that tends to be inherited together) with separate neighbouring genes that may have a behavioural effect. Why are slow COMT prepubertal girls taller, for example?

As mentioned above, we know of COMT from Parkinson’s disease (PD) treatment. One might expect high COMT people (not patients!) to have parkinsonian like motor or cognitive deficits, and low COMT to have excess movement and anxiety. Note however that the only clinical COMT inhibitor that acted centrally was tolcapone, which has been virtually withdrawn because of serious liver side effects. All entacapone and opicapone do is act peripherally to make orally administered L-DOPA stay in the blood stream for longer.

The purpose of this journal club is to go back to the science on the topic. To try and gain further understanding of how the polymorphism is changing the neurochemistry of the brain rather than just looking at the end result expressed as human behaviour, a transgenic mouse model of fast COMT was developed. This was specific for cortex, striatum and hippocampus COMT and not other sites like the cerebellum and outside the brain. They confirmed that the Val158 allele polymorphism had an equivalent human level 30% level of overactivity of COMT and found that the mice had deficits in stimulus response learning and working memory similar to those in human studies. They also showed signs of impulsive and compulsive behaviour. These are the kind of deficits expected to be related to dopamine signalling. They had no generalised cognitive or motor deficits, which is interesting in relation to PD.

They didn’t record reduced synaptic dopamine levels because apparently dopamine is difficult to measure in mice, but they electrically stimulated anaesthetised transgenic mice and did find increased dopamine release compared to controls, that they suggested was compensatory. One might also expect a compensatory increase in dopamine synthesis on other measures, eg increased tyrosine hydrolase (the enzyme that produces dopamine) or reduced dopamine transporter activity in the striatum (dopamine transporter clears dopamine back into presynaptic terminals to reduce the dopamine signal), but they did not find this. Strangely, the rate of dopamine decay was unchanged. So if there is no demonstrated change in dopamine activity in the synapse, what exactly are these changes compensating for? It was suggested that COMT was influencing dopamine regulation presynaptically rather than affecting breakdown of synaptic dopamine.

Given this somewhat confusing information, indicating only that there is at least one more layer of complexity to COMT, it is worth going back to more recent scientific literature to see what magnitudes of human differences are actually present between the different genotypes.

Scientific reports state that the enzyme is 25-33% less active (Tunbridge 2019) not 25% as active in methionine homozygotes. These assays are simply in peripheral blood; a functional MRI study could only measure differences in general brain activity in different genotypes while radioactive ligand PET or SPECT imaging is not looking at dopamine levels or COMT levels but again just differences in regional brain metabolism or blood flow. A DAT scan is just measuring the structural integrity of presynaptic terminals in the striatum. It takes a 90% reduction in these terminals to give clinical PD so one could not hope to see some compensatory change in DAT transporter levels from mild changes in COMT activity.

The subject of this journal club is a cognitive study by Zareyan S, Zhang H, Wang J, et al. First demonstration of double dissociation between COMT-Met158 and COMT-Val158 cognitive performance when stressed and when calmer. Cerebral Cortex. 2021;31(3):1411-1426.

I wanted to avoid yet another metanalysis and focus on individual study design. The first thing the paper clarifies is that there is relatively better executive function in Met subjects, despite the scare-mongering from the “commercial” literature as described above, and Met homozygotes are not necessarily worse. But even these findings are not universal. They sought to test a specific hypothesis, namely the prediction that the effect of stress on cognitive performance would specifically raise the performance of Val homozygotes because the stress induced extra dopamine would allow them to compensate for their background dopamine deficiency, and the performance of Met individuals would decrease under stress conditions under the same protocols.

In 140 subjects, having excluded those with pre-existing stress disorder, the stress was to have two testers in the room at the time standing behind with a clipboard. They measured salivary cortisol level to check baseline stress level and made careful efforts to make this baseline as low as possible.

The cognitive tasks included a Flanker Test, where attention is directed to a central visual stimulus on a screen and the subject is to ignore flanking stimuli. The central stimulus points left or right and the subject has to press the correct button as quickly as possible. Then the situation is reversed and the subject must ignore the central stimulus. There was also a reasoning intelligence test and a Hearts and Flowers task, where the stimulus lights up to left or right and the subject has to press the button corresponding to the opposite side.

Unfortunately, on their genetic testing, only 7% were homozygous Met met instead of expected 25% so they mainly compared homozygous Val with heterozygous Met.

The mean reaction times went from 630 ms to 600 ms under stress for Val homozygotes (ie better), and from 610 to 630 ms for Met carriers. There were similar level changes in the other measures. While significant, these differences are not exactly life-changing. From previous similar studies, the differences for Met homozygotes were not much more striking. The study confirmed that there was no speed-accuracy trade off.

The study discussed how other studies might have missed these modest effects. It might have to be exactly the right level of mild stress to make a difference ot otherwise all subjects would perform worse. Alternatively the stressor might have to be task specific or relate to a certain kind of attention demanding task. For example, in Huntington’s disease, a condition where there is severe striatal and prefrontal dysfunction in some ways opposite to those in PD, there are specific deficits with anti-saccades, an attention demanding task that has similarities with the tasks performed in this study.

The study also considered that there could be a ceiling effect where those with better baseline working capacity have less room to improve and therefore more likely to get worse with distraction or stress. They commented on the related transmitter noradrenaline (NA) and that an intermediate level of COMT activity may similarly be optimal, but that there was little evidence that COMT polymorphisms actually affected peripheral NA levels.

Overall, this seemed to be a carefully conducted study with a clear hypothesis and a clear result, despite the bad luck of not recruiting many Met homozygotes.

Turning to relevance for the general population, it is completely unclear if these measured differences have any practical meaning, enough to label someone as “deficient”, “handicapped” or “divergent”. This is very hard to justify anyway when 50% of the population is homozygote one way or the other. These are statistical effects measured across many subjects; there will be huge variations between individuals based on many other genetic and environment factors. Many more scientific studies are needed on the magnitude of these effects in practice. There are also potentially unpleasant connotations of “branding” individuals, or even worse whole populations, with a personality type based on a genetic polymorphism. Are such genetic tests going to be part of job interviews or criteria for life insurance or health insurance policies?

One potentially more helpful connotation may be sensitivity to drug side effects, specifically drugs that act on the dopaminergic and noradrenergic systems such as SSRIs, SNRIs and monoamine oxidase inhibitors in patients with or without an additional dopaminergic disease deficit. For example, might tyramine reactions be worse in slow COMT individuals because of increased lingering of dopamine and NA at the synapse. Or are the effects very complex and unpredictable as suggested by the mouse studies?

Certainly, testing for susceptibility to drug side effects is not unheard of; TPMT activity is performed routinely before giving azathioprine and one might test for certain genetic congenital myopathies before giving certain anaesthetics. What would be needed is an actual study looking at the incidence of tyramine reactions, neuroleptic malignant syndrome and other less severe drug side effects in different COMT genotypes and, if a major difference was found, genetic testing beforehand would seem to be a very valid and cost effective strategy.

Posted in Parkinson's Disease, Psychiatry | Tagged , , | Leave a comment

Risk of Further Seizures after an Unprovoked Seizure

Posted on September 8, 2026 by dulcetware

Epileptic seizures are a common occurrence, with a lifetime prevalence of around 5%, and around two-thirds of these being unprovoked seizures, namely not a manifestation of some underlying pathology such as a metabolic insult, brain injury or structural brain lesion. Uncertainty remains over whether one should treat an unprovoked seizure with the aim of preventing a recurrent seizure.

This subject was discussed in a previous journal post, but since it was based on a study from around 2006 and another from the 1990’s, it seemed worthwhile to revisit the topic in the light of recent Cochrane reviews. It is interesting to compare our “wish-list” compiled in 2013 of what a more definitive study would look like, versus what has actually happened in the intervening decade.

The main issue remains the same: if we don’t know what the recurrence risk is after a single unprovoked seizure, and the level of reduction of risk by taking antiepileptic drugs (AEDs), how can patients make an informed choice on whether to embark on such long term treatment?

I will first discuss a related Cochane review that looked at risk factors for a recurrent seizure as it highlights the limitations of pooling information from disparate studies. This review took 5918 patients across 23 studies; only one was an RCT and in only seven were AEDs investigated. A major issue was that not even the statistical calcuation was standardised; some used relative risk, sone odd ratios and some hazard ratios. While broadly measuring the same thing, they can give widely different values and it is worth for the sake of this journal club about critically appraising studies to explain what the measures are.

Risk, statistically, is the number of times an event occurred divided by the number of times it could have occurred, so 1/6 chance to get a 1 on a six-sided die.

Odds is the number of times an event occurred divided by the number of times it didn’t occur, so 1/5 in the same example. The difference is not quibbling when the risk is high; a 0.5 risk becomes 1.0 odds.

Relative risk is simply one risk value divided by another, and the odds ratio is analogous.

There are not hard and fast rules but, generally, an odds ratio should be used in a case-control study, with two separate populations (the haves versus the have-nots, or the “1”s on a die roll versus the “non-1″s).  A classic example is the proportion of smokers in lung cancer cases versus the proportion of smokers in non cancer controls. Performing linear regression gives an odd ratio result.

Relative risk should be used for a cohort study, where there is one population and some are exposed to some factor, or in the case of a randomised controlled trial, the exposure is a controlled intervention. An equivalent example would be taking a cohort and following up if they developed lung cancer while seeing if they had started smoking; the calculations would be the rate of cancer in the smokers versus the overall rate, and the rate in the non-smokers versus the overall rate.

An aside, with which this journal club is relatively or absolutely obsessed, is the difference between absolute differences versus relative differences. An intervention that changes a risk from 2% to 1% could be quoted as 50% improvement or 1% improvement depending on what case the authors wish to make.

The statistical hazard specifically refers to a time to event analysis, namely cox proportional hazard regression analysis. A typical Kaplan- Meier plot can give a relative risk looking at the difference in survival rates at a certain time point, when the start values were obviously 100%. The hazard ratio is the instantaneous risk at a certain time point given they had already survived to that point, not the total risk from time zero to that point. It is more like the slope of the graph at that point, in fact specifically the slope of the graph at that point divided by the survival probability at that point.

So the Cochrane review produced a mix of relative risks, hazard rations and odd ratios. For example family history in a first degree relative gave a relative risk of 1.47 (1.16 to 1.85 confidence interval) and a hazard ratio of 1.35 (1.03 to 1.76). The only other ones that reached significance were abnormal imaging, abnormal EEG, Todds paresis (just), and noctural seizures. Perhaps surprisingly, status at presentation was not a risk. Child versus adult age gave conflicting significant results for odds ratio versus relative risk in different studies.

That all being said, the main subject of this journal club, Prognosis of adults and children following a first unprovoked seizure by Neligan et al (2023) had even more patient numbers – 12160 across 58 studies, most being cohort studies (relative risks anyone?)

They included status and a cluster within 24 hours, as standard, and usually excluded myoclonic jerks and infantile spasms as they are rather obviously part of ongoing epileptic syndromes. In adults the 6 month risk was 25%, at 12 months 35%, at 24 months 41%. These closely match the rates of the FIRST study of 1993 from the previous journal club and higher than the subsequent MESS study.

The study looked at a number of other parameters, but these were more based on individual studies and did not necessarily show the effect of AEDs. Delaying treatment in the FIRST study until after second seizure did not increase mortality eg at 5 years, 97% (95 to 99%) in immediate versus 98% (95 to 100%) delayed, but there is more to treatment than avoiding mortality and it simply shows that the risk of SUDEP is thankfully low and so it is difficult to tease out differences.

High risk subgroups were as expected: traumatic brain injury gave 82% risk at 12 months if the first seizure was after 7 days following the injury, and there was a high risk after stroke. There was a somewhat higher risk in patients with Alzheimer’s dementia and lower in Lewy body dementia.

In the very long term NGPSE study, if a patient was seizure free for 5 years after single seizure, their risk of subsequent seizure at 10 years was still 10%, but if free for 10 years, the risk at 20 years was 1%. There are not clear data on the effects of AEDs on this risk.

In conclusion, the Cochrane reviews support the 40% 2 year risk quoted figure for seizure recurrence after a first unprovoked seizure, but there is no clearer information on treatment strategies. It would take a study with the design of the wish list of the previous journal club to answer those questions rather than another Cochrane review pooling existing data.

 

Posted in Epilepsy | Tagged , , | Leave a comment

Tofersen in Motor Neurone Disease (Amotrophic Lateral Sclerosis)

Posted on June 21, 2026 by dulcetware
The current image has no alternative text. The file name is: tofersen_front_page.jpg

Motoneurone disease (MND), called Amyotrophic Lateral Sclerosis (ALS) in the US, is without doubt a devastating neurodegenerative disease. The standard pharmacological treatment for the condition is riluzole, which results in a modest but definite slowing of disease progression. Since its inception in the 1990’s, many studies have confirmed its efficacy, showing that it slows the deterioration in quality of life rather than just prolonging survival. It costs around $50 to $200 dollars a month.

Unfortunately, over the subsequent 30 years, no better treatment has been identified. However, there have been recent advances in gene therapy that may have a bearing on the condition.

About 2% of all cases of MND are due to a mutation in the SOD1 gene. Most cases are sporadic and not as readily amenable to gene therapy. Tofersen is an antisense oligonucleotide that binds to the SOD1 mutation gene. It is somehow targeted by cell membrane receptors that trigger endocytosis. Then there is leakage from the endosome into the cytosol where it binds to SOD1 mRNA to prevent translation; RNAase recognises it as a hybrid and destroys it. The result is that it downregulates the protein product of the SOD1 gene by 80-90%. As an autosomal dominant mutation, no protein is better than mutated protein. SOD is one of three antioxidant enzymes for clearing superoxide and converting atomic oxygen into hydrogen peroxide; presumably the other two can do the job.

The drug has to be delivered into the cerebrospinal fluid to work. This entails a course of fortnightly intrathecal (lumbar puncture) injections for 6 weeks followed by monthly injections. It is not clear if this is to continue indefinitely.

To obtain the treatment privately, including the infrastructure required for intrathecal delivery, has been estimated at €450,000 a year! While the drug is in a trial stage, the drug company Biogen are offering the drug itself for free.

In our journal club, we discussed preventative therapy; we found that the penetrance of pathogenic mutations is significantly variable and averages a little over 50%, though some mutations have much higher penetrance. We are not going to deliver intrathecal injections monthly when someone might not ever get the disease.

The initial phase 3 trial of tofersen, which they called VALOR, had 108 patients enrolled and was completed by 97. Two thirds of the patients had treatment, the remainder placebo. It ran for 28 weeks from 2019-2021 and did not reach significant improvement at end point.

The topic of this journal club is an open label follow up of the same patients, which came to the opposite conclusion.

It is worth going into more detail on the original study. This primary end point was the score on the ALS functional rating scale in a subgroup of patients predicted to have faster disease progression, presumably so that they could see an earlier difference from treatment.

The range of this scale is 0 to 48 and after 28 weeks the deterioration in the treated group was 7 points, and in the untreated group 8 points; the 95% confidence interval of the difference was -3.2 (drug worse) to 5.5 (drug better), p=0.97. This lack of significance could hardly be clearer and moreover one point difference on a 48 point scale does not sound like a meaningful benefit even if it was significant.

The 95% CI for vital capacity difference was -3.5% to +19 % while hand held dynamometry strength testing was negligibly different. The rates of death or permanent ventilation in the two groups was 2/21 on placebo and 4/39 on treatment.

They described LP related adverse events as common and serious adverse events occurred in 7% of tofersen recipients. These were mainly myelitis (inflammation of the spinal cord), radiculitis and papilloedema.

In this paper, they also discussed one year data, which included a 6-month open label extension period, meaning that both originally treatment patients and the placebo patients now received treatment. These data were pooled over all the patients, not just the faster progressing ones. They suggested that there might be delayed benefit, but this did not change the study’s overall conclusion.

They didn’t go into detail on the combined results for rapid deteriorators and slow deteriorators at 28 weeks, the end of the placebo period, because they said they didn’t plan to adjust the statistics for multiple comparisons, but on looking at their graph, it appears to be about -4 points deterioration on the scale in the treated patients versus -6.25 for untreated. The % change in vital capacity looked significantly worse in untreated, and there was no significant difference in hand dynamometry, and no significant difference in rates of death or requiring permanent ventilation.

The topic for this journal club was a longer open label extension up to 3.5 years. Of the original 108 patients, 94 completed the initial 28 week study, but only 46 completed the open label extension!

This should send one immediately diving into why this was the case. By the end, there were 34 patients still left in the early start group, and 12 in the delayed start group (the original placebo group). The original numbers were 63 and 32.

In the delayed start group, 21 discontinued – 1 adverse event, 6 withdrawn consent, 6 death, 5 disease progression, 3 other.

In the early start group, 29 discontinued – 1 adverse event, 1 lost to follow up, 4 consent withdrawn, 14 death, 8 disease progression, 1 other.

So more patients stopped due to death in the early treatment group (22.2% versus 18.75%).

There was reduced csf SOD1 protein in the treatment group, and neurofilament light chain, but the latter equalled out by the end of the trial. It is often wishful thinking to use biomarkers as a surrogate for clinical outcome. They might be more useful in normal clinical practice to monitor individual patients.

The early start group (both predicted rapid deteriorators and slow deteriorators) had 9.9 points deterioration on the functional scale versus 13.5 points, and -13.8% reduced vital capacity versus -18.1%. The authors did not highlight the statistical results, saying the study was not powered to show this, but referred to “numerically less decline” in the early treatment group. Clearly, they were not significantly different. The p value for functional score was 0.14, that for vital capacity was 0.44 and dynamometry 0.55. Indeed, on the graphs, the vital capacity figures appeared to come together by the end of the trial and the dynamometry figures never really separated in the first place.

Regarding survival (avoiding death or permanent ventilator dependence), the hazard ratio was 0.64 (0.28 to 1.46 95% CI), so clearly there was no evidence of shortening of life in the delayed treatment group. On the Kaplan Meier survival plots, there were some strange findings. Overall, there was very little “numerical” (to use the authors’ terminology) difference at 28 weeks or 3.5 years, but between those times the delayed start had a worse trend. In predicted milder patients, early starters had a major worsening by the end of the plot, and in predicted severe patients, there was initially no difference (as reported in the first study), but when the delayed start patients later got treatment, they fell off more rapidly instead of catching up! This is probably explained by the tiny numbers of patients being analysed.

Statistical adjustments were apparently made for riluzole use, which might have varied between groups. We did not go into detail on what these were, but clearly a treatment which is of proven benefit might confound a study on one which is of no proven benefit when originally compared with placebo.

So, in essentially the same trial, the authors report no difference between treated patients and untreated patients, and that there were significant adverse events, but when the untreated patients later receive treatment, they say they are worse off after 3.5 years (“numerically” worse not statistically worse).

The stated purpose of the study was to compare early treated versus delayed treated patients. Yet the authors reported that the study, conducted over years in many centres across a number of countries, was not powered to show differences between groups. There appeared to be no clear hypothesis beyond this. Normally, in a delayed start protocol, the hypothesis is that there is a symptomatic benefit and disease modification benefit and if the delayed start group never catches up, it is presumed there was originally some enduring neuroprotection. It is of course difficult to test such a hypothesis when the treatment shows no symptomatic benefit in the first place.

Instead, the authors projected a comparison with what they estimated in other studies to be the natural history when untreated. This is not a randomised placebo controlled study but an observation on two completely different uncontrolled samples collected in different studies.

The conclusion was the opposite of their own initial VALOR study; now, by comparing two groups, both of which have been receiving the same treatment for years, they have “demonstrated the benefit of toferson in SOD1 ALS and a clear rationale for its use“.

The reason for these apparent contradictions, which the paper glosses over, seems to be that the benefits of treatment may have been in the opposite group to that they expected, namely in milder cases. As expected with delayed treatment, the two groups then converged, though there was a non statistically significant suggestion of maintained advantage in the functional scale in milder patients who had earlier treatment. There was certainly no benefit in death/ permanent ventilation.

The UK MND association, who stated they had met with Biogen the drug company, reported on 17th March 2025, “Tofersen gives not just a glimmer of hope to people with SOD1 MND, but a future. It is the difference between life and death. The drug itself costs nothing at the moment – yet we risk people dying for want of the NHS not being able to pay for the cost of administering it.”

The inference drawn here is different. A study was conducted on a rare subgroup of MND patients. To achieve a quick result they focused on a subgroup of patients predicted to show rapid deterioration. They seemingly could not recruit enough patients across multiple centres in many countries. There could not have been a clearer negative result (p=0.97). Ethical approval was probably contingent on the placebo arm getting the treatment later, and so there was an open label extension regardless of the results of the study. Pooling the data from the minority of patients who continued to have the intrathecal injections and who were still alive (there was a greater proportion of drop out due to death in the early treatment group), there was a non significant trend in one of the measures for a benefit in patients who had started earlier.

Treatment for such severe conditions is inevitably emotive, especially when there have been no pharmacological advances for 30 years. But that does not constitute a waiver in respect of permission to rely upon “numerical” differences and comparison with natural history in other populations. For end of life conditions that typically carry at least as devastating a prognosis as that of MND, the NHS applies a higher limit of QALY (Quality Adjusted Life Years) of £50,000; this is the cost of an extra year of perfect health that can be afforded. Even relying upon numerical non-significant results, the stark facts are that the placebo controlled section of this study showed a 1 point improvement in a 48 point scale and cost €450,000 a year and requiring 12+ lumbar punctures. The open-label extension that showed a “clear rationale for use” revealed about 1 point improvement per year over around 3 years.

We all desperately hope for effective treatment for neurodegenerative conditions such as MND. There is a risk that the sheer cost and infrastructure involved in treating this 2% of patients with MND with tofersen is a distraction from finding such treatment.

Posted in Uncategorized | Tagged , , | Leave a comment

Incidence of Sudden Unexpected Death in Epilepsy

Posted on April 24, 2026 by dulcetware

Seizures are a common condition, experienced by around 10% of people in their lifetime, and while in some cases epileptic seizures may be the consequence of a particular brain insult, a significant proportion have an ongoing susceptibility to epilepsy, rather parsimoniously defined as having had at least two unprovoked epileptic seizures. The distinction between a single unprovoked seizure and epilepsy is somewhat moot, since around 40% of those who have such an event will have another within two years if left untreated.

Perhaps unhelpful, or harmful, would be a better description than moot. Seizures are not a trivial condition; quite apart from the shocking nature of the event and the restrictions that it imposes upon lifestyle, such as the legal entitlement to drive a car, they are associated with a small but real mortality risk. While this is more likely in those who have frequent seizures, those who have had a single seizure are not immune.

The mortality asociated with epilepsy is termed sudden unexpected death in epilepsy (SUDEP) and was defined originally by Nashef as:

“a sudden, unexpected, witnessed or unwitnessed, non-traumatic, and non-drowning death in an individual with epilepsy, occurring during benign circumstances, with or without evidence of a seizure, excluding documented status epilepticus, where postmortem examination does not reveal a cause of death.”

Note that SUDEP is not about experiencing a seizure in a dangerous environment in which to lose consciousness or control, or even necessarily due to having a seizure at the time of death, though this is generally assumed.

Individuals who have epilepsy are often young and otherwise well and so SUDEP is obviously an extremely emotive topic, both for sufferers and doctors. There are cases where neurologists have been sued for not trying hard enough to make their patients compliant with medication and not informing them of the risk of SUDEP.

But how do we inform them of the risk if we do not know? In developed countries the risk in studies varies from 0.35 to 9.3 /1000 patients with convulsive epilepsy per year – a thirty fold variation. Such a wildly varying estimates makes a meta-analysis proclaiming that they average out at 1.2/1000 practically useless.

It remains unclear if the inaccuracy of SUDEP risk is because of varying definition, underreporting, no post mortem to prove that there is no other cause, population differences or a combination of factors. The studies that have been conducted have usually been on an uncontrolled sample and retrospective.

As well as informing patients of the risk, if one understands the level of risk, one might ascertain risk factors and so be able to mitigate them by medical intervention or education.

Again the findings are inconsistent between studies and range from male sex, suffering generalised tonic clonic seizures and poor medical complaince to seizure frequency, young age of onset, sleeping alone or being on multiple antiepileptic drugs.

The topic of this journal club is a large cohort long term prospective study published last year in Seizure: European Journal of Epilepsy by Rongxin Li et al, Incidence and risk factors of sudden unexpected death in epilepsy in rural Northeast China.

They recorded Nashef’s definition of definite, probable and possible SUDEP in a prospective epidemiological cohort of patients with convulsive epilepsy, conducting face to face interviews for epilepsy with monthly assessments of a median of 7 years. They had 6513 patients who had had a median duration of epilepsy of 32 years. 855 were lost to follow up. This meant there was a total of around 44,000 patient years. All were treated with valproate or phenobarbitone, or additional AEDs prescribed by a specialist.

The headline finding was 64 SUDEP deaths, which gives an annual incidence of 1.4/1000.

Regarding risk factors, they found the following statistically significant risk factors on univariate cox regression:

  • Higher BMI p=0.0116
  • Earlier age at onset p = 0.001
  • Longer duration of epilepsy p = 0.007
  • Higher generalised seizure frequency in last month or last year p=0.001

There were in fact 9 times as many non SUDEP as SUDEP deaths in this cohort, so they could use these events as competing risks to reduce bias. Using multivariable competing risks regression on the factors that were significant on univariate analysis, they determined relative risks:

  • Obesity had relative risk 1.06 (95% confidence interval:1.02 to 1.11)
  • Later age of onset was a protective factor 
  • Increased seizure frequency in last month had relative risk 12.10 (95% CI 7.39-19.84)
  • Increased seizure frequency last year had relative risk 11.56 95% CI 1.59-83.92)

Comments

It is worth putting the 1.4/1000 figure, which closely matches the consensus metanalysis figure of 1.2/1000, in the context of overall risk, as many patients may not grasp their significance. The risk of accidental death in the UK is 20,000 per year. Out of 68 million, this gives 1 in 3000, or 0.3/1000. So the risk of having epilepsy is four times as great. For further comparison, patients often want brain scans as they are worried they might have a brain tumour. Its incidence in the general population (and likely also in the population of people with chronic episodic headache) is 20 per 100,000 per year, so 1 in 5000, or 0.2/1000.

Regarding life time risk, that of death by RTA is 1 in 100, but the lifetime risk of SUDEP in a patient with epilepsy has been estimated at 7-12%. (This seems quite high; if a patient has epilepsy for as long as 50 years, this is 1.2 * 50 = 6%.)

The criteria for probable SUDEP are as follows: a history of epilepsy in the patient; sudden and unexpected death while in a healthy state; death occurring within minutes; death occurring during normal activities in benign circumstances; and no definitive cause of death identified upon clinical examination.

The criteria for possible SUDEP are as follows: the presence of a competing cause of death; no other significant underlying disease; death occurring in water but without complete submersion of the individual; and death occurring in a normal environment without an autopsy.

But, having explained all this, the study ignored it and lumped all possible and probable cases with definite cases because they had no post mortem data or examinations at death. This is a real issue; in a previous study in Chalfont in 2013, 40% of the apparent SUDEP deaths were cardiovascular related and not epilepsy related. The death rate not from SUDEP in the study’s population was quite high, increasing the possibility of uncertainty over cause.

In addition, the study explained that there was no correlation with AED adherence, but they only asked at assessments. One would have to question if patients recruited in a single cohort followed every month for many years would be reliable in retrospectively reporting their compliance. Other studies have recorded AED drug levels at autopsy and have indeed found a correlation, as might be supposed.

While the study might overestimate risks somewhat because of lack of post-mortem data, the prospective data ascertainment would seem to give a reliable figure. Moreover, the monthly assessments could give good data on seizure frequency, which was found to be the major risk factor for SUDEP.

Finally, as the study points out, it was conducted in a single rural region of likely uniform population and ethnicity. This might be somewhat differnt from the population in other countries.

Conclusion

The study gives more confidence in citing the figure of around 1.2/1000 per year for the risk of SUDEP. Of particular interest was the major risk factor of increased seizure frequency; in clinical practice, if a patient reports increased generalised tonic clonic seizure frequency, they should be assessed as a matter of some urgency rather than waiting for routine follow up, which in the UK may be six months or a year.

Posted in Epilepsy | Tagged , , | Leave a comment

Endovascular Treatment of Cerebral Venous Sinus Thrombosis

Posted on December 4, 2024 by dulcetware

Introduction

Cerebral venous sinus thrombosis is blockage of one or a number of venous vessels, often called sinuses when they are large vessels, that drain blood from the brain. This results in back-pressure, oedema and ischaemia that develops rather gradually compared with an equivalent blockage of an artery.

When severe, the ischaemia can result in infarction, i.e. permanent damage to the brain tissue. The back-pressure may result in general raised intracranial pressure, which in turn causes other effects including local haemorrhage, headache resulting from stretching of the meninges and other structures, interference with autoregulation of cerebral blood flow and papilloedema which, if left unchecked, can leave lasting retinal damage.

It might immediately be inferred from these facts that venous stroke, i.e. brain infarction, presents differently from arterial stroke; it is generally more gradual in evolution, with a wider range of manifestations. It is also much more rare. All these factors make it considerably harder to diagnose.

Another difference between arterial and venous stroke is the pattern of focal neurological deficits stemming from differences in blood supply. The cerebral veins are typically central, draining both sides of the brain, so the deficits are correspondingly bilateral rather than contralateral.

The superior sagittal sinus runs posteriorly over the convexity of the hemispheres along the central sulcus (this is why neurosurgeons tend to “steer clear of the midline”) and joins with the straight sinus to split into the two transverse sinuses. These also receive drainage from the cerebellar hemispheres, brainstem and veins directly supplying some cranial nerves. The transverse sinuses run into the sigmoid sinuses which then leave the skull as the internal jugular veins. Thus, sagittal sinus thrombosis may result in unilateral or bilateral cortical hemisphere deficits or in seizures, while transverse sinus thrombosis is more likely to result in general obtundation. Sometimes one transverse sinus is much more dominant than the other, making even these paired structures potentially likely to result in bilateral deficits if thrombosed.

In parallel, the deeper structures of the hemispheres are drained by the inferior sagittal sinus running deep within the central sulcus. This joins the great vein of Galen to form the straight sinus. The great vein of Galen drains the diencephalon structures and internal capsule and thrombus here is again likely to result in obtundation and behaviour changes as well as potentially bilateral long tract signs.

Finally, the cavernous sinuses behind the eyes have a number of cranial nerves and the carotid arteries running within them. These drain posteriorly via the petrosal sinuses into the sigmoid sinuses.

In contrast to the evolution over hours of arterial thrombosis, the symptoms of venous thrombosis evolve over days and diagnosis is often delayed by several days as a result. A plain CT head will detect up to 2/3 of sinus thrombosis from the hyperdensity of the posterior sagittal sinus near the transverse sinuses. Subarachnoid and parenchymal haemorrhage is also common and may be picked up on CT.

A CT venogram is more than 80% sensitive in detecting venous thrombosis, but due to false positives of normal anatomy is less than 50% specific.

MRI head may only pick up venous thrombosis after several days; complex signal changes take place on T1 and T2 due to methaemoglobin changes over this time while oxyhaemoglobin changes over the first few hours have unpredictable timing. In general, gradient echo is the best sequence. Regarding ischaemic and infarction changes, venous stroke often spares the cortical rim due to oedema building up more in the deeper layer of the grey-white matter junction.

MR venography is a sensitive non invasive investigation specifically for venous thrombosis.

If diagnosis of venous sinus thrombosis is difficult, management is still more uncertain. The American Stroke Association recommends anticoagulation for around 3 months or until demonstrated recanalization, even if there is secondary haemorrhage, but this is based upon very flimsy evidence. A Berlin trial around 25 years ago using unfractionated heparin showed some benefit, while a small Dutch trial in 1999 did not, though those patients may have been less severely affected on average.

It seems an attractive proposition to engage in more direct treatment of a clot, mechanically removing it or giving anticoagulation/ thrombolysis where it is needed, with less risk of worsening haemorrhage in distal or remote brain tissues.

However, it would not be the first time that a treatment might sound reasonable and yet not be efficacious. The only randomised controlled study before 2020 on thrombectomy (mechanically removing a clot) was terminated for futility as it was clearly showing no advantage over standard systemic anticoagulation.

The Current Study

It was against this background that the current study, Endovascular Mechanical Thrombectomy and On-Site Chemical Thrombolysis for Severe Venous Sinus Thrombosis (Liao et al., 2020; nature Scientific Reports, 10:4937) was reported. They essentially reviewed their consecutive cases between 2005 and 2015. Of 30 cases (compare that with arterial stroke!), 16 remained stable or improved and so had conservative therapy while 14 deteriorated and had intervention.

This intervention consisted of a catheter inserted up the venous circulation, manual aspiration of thrombus under angiographic guidance and bathing the vein in a thrombolytic again via a catheter for around 24 hours.

They found that, despite being a more severely clinically (and radiologically) affected group of patients, the eventual outcome as measured by modified Rankin score ( the same score used for arterial stroke) was not significantly worse than in the group that stabilised spontaneously on heparin. Specifically, 75% of the conservative group had a score of zero deficit on discharge, compared with 64.3% in the intervention group (p=0.504). At three months, the conservative group had 87.5% with a score of 0-1 and the intervention group 85.7%. I saw no breakdown of no versus mild deficit at three months, but the 0-1 figures on discharge were almost identical at around 85%, so return of function was rapid. In fact, the 2 remaining non 0-1 score patients on intervention were dead, while 1 patient in the conservative group had a score of 4 and one had a score of 5.

The authors concluded that intervention was non-inferior and relatively safe (the two deaths were not directly related to the procedure – one had intracranial haemorrhage progression (possibly related to thrombolysis depending on where in the brain it was) and the other had a pulmonary embolus). One can presume that, since the intervention patients started off worse rather than being randomised, this argued in favour of intervention in such severely affected cases. As is obligatory in every discussion, they suggested further studies.

They also compared results with another study done at around the same time. This was the TO-ACT study which was actually prospective and randomised 67 patients to intervention vs anticoagulation. At 12 months 67% of the intervention group and 68% of the heparin group had a good modified Rankin score of 0-1. Mortality was 12% in the intervention group vs 3% in the heparin group (p=0.2) and symptomatic intracerebral haemorrhage was 3% in the intervention group and 9% in the heparin group. The conclusion was obviously that there was no better outcome. Note that the extra deaths form part of the overall Rankin figures so it is a fair comparison, a little like the haemorrhage data for intravenous thrombolysis in arterial stroke where some patients did worse but the majority did better. Nevertheless, if you tell a patient there is 12% mortality if you perform thrombectomy and 3% if you don’t, not many are going to want to go ahead even if you plead it is not statistically significant at p=0.2.

Study Appraisal

Some interesting facts came out of the demographic data. The range of time from onset to diagnosis in the stable patients was 2 to 13 days, with a median of 7 days – clearly it is banal to make a direct comparison with time course and management of arterial stroke.

Other than this, the numbers are far too small to draw any conclusions at all. We find ourselves drilling down to individual patients to account for the Rankin scores in the two groups.

The statistical conclusion can only be that there is no evidence to show that thrombectomy has a different outcome from heparin therapy. Time and again, we must remind ourselves, and the journal Nature it seems, that this does not equate to evidence for no difference. Once more, no evidence for a difference is not the same as evidence for no difference. The latter requires a power calculation and it should have been possible to predict from the 2012 study that one was going to require thousands of patients for that, not 15 in each arm. The 67 total patients in the other study was equally doomed; perhaps there would be an advantage apparent in enough patients, or in the severely affected patients.

As a case series, one can infer that the intervention in their centre was a reasonable treatment option. The patients did really quite well despite being in a serious and deteriorating clinical category. The numbers are too small to assess safety more definitively.

Given the duration of their study and the multiple centres of the TO-ACT study that was only able to recruit 67 cases, the sad conclusion is that, in this rare disease, we are never likely to have the benefit of a watertight randomised study investigating interventions in appropriate clinical subgroups.

It may be one of the many clinical conditions where we must rely on clinical judgement and experience in individual cases, with the reasonable inference that, if the patient is otherwise relentlessly deteriorating, mechanical thrombectomy, at least in experienced hands, may offer a reasonable balance of risk versus benefit.

Posted in Stroke | Tagged , , , , , , | Leave a comment

Inhaled Levodopa

Posted on March 24, 2023 by dulcetware

Introduction

Levodopa was the first main, and arguably still the best, treatment for Parkinson’s disease. Being similar in structure to tyrosine, it is absorbed through the gut by an amino acid transporter – its half life in the blood is short, namely 0.7 to 1.4 hours. This has advantages in terms of avoiding peripheral dopaminergic effects and dopa decarboxylase inhibitors further reduce this latter effect.

The fact that the clinical action of levodopa is far longer than its half life is presumably due to presynaptic storage and control of release as dopamine and so it is no wonder that, as the disease progresses and there is increasing loss of the remaining presynaptic terminals, PD control from oral levodopa becomes more brittle and the clinician resorts to frequent dosing, longer acting agonists, delayed release preparations and even continuous intrajejunal delivery.

This review article, Profile of inhaled levodopa and its potential in the treatment of Parkinson’s disease, describes a further new delivery method, namely inhaled levodopa. This avoids the variability of gastric emptying and competing with the amino acids of food for absorption.

CVT301 inhaled levodopa is a complex of molecules in a large low density porous particle that can aerosolise and avoid phagocytic destruction in the lungs. The particles are readily absorbed through the alveolar membranes. The Arcus inhaler works by delivering the particles in dry powder form and requires the patient only to breathe in rather than also to activate a pump.

Phase I study

In a Phase I study, following safety trials in animals, plasma levodopa concentrations rose within 10 minutes of inhalation (they also took oral carbidopa) by healthy volunteers.

Phase II Study

Ini this study, patients with Parkinson’s disease and significant off periods had a median T max (time to max concentration ) of 15 minutes vs 66 minutes for oral, and clinical improvement (finger tapping) within 5 minutes and lasting 90-100 minutes.

Side effect of cough occurred in 25% which often settled after initial dose and was described as mild. There was less dyskinesia giving 50 mg than for oral 100 mg – but the study did not quantify the relative benefits of finger tapping tasks, so perhaps the oral was simply a bigger dose with more benefit and more side effects.

There were no differences in lung function between oral and inhaled groups after 4 weeks.

Phase III Study

There was a 12 week double blind, placebo controlled study in 339 patients with motor fluctuations randomised to placebo, 35 mg dose and 50 mg dose. Chronic lung disease was a contraindication.

Mean change on motor UPDRS at 12 weeks from pre treatment versus 30 minutes after treatment was -9.83 versus -5.91 for placebo (p = 0.009). There was no reduction in off time according to patient diaries; 85% of subjects completed the study. There were side effects of cough (14.5% versus1.9% placebo), URTI (6.1% v 2.7%), nausea (5.3% v 2.7%), sputum discolouration (5.3% v 0%), and dyskinesia (3.5% v 0%). Only two subjects discontinued due to cough. No statistically significant differences in FEV1 or diffusion capacity of lungs were found (but is this evidence for no effect on the lungs rather than no evidence for effect?).

Long-term studies

In a longer term open label study of 408 subjects, the treatment arm used a mean of 2.3 doses a day. After dosings at four weeks into the study, they experienced a mean UPDRS III change of -5.7 at 10 min, -12.0 at 20 min, -15.5 at 30 min and -16.1 at 60 min at 4 weeks and at 60 weeks the equivalent changes were -5.0, -11.5, -15.3, -14.8.

In another study, FEV1 and diffusion of carbon monoxide were not statistically different (!). Dyskinesia was worse; 5.5% versus 3.1% in the group with standard of care. Two of I think 100 subjects discontinued due to cough.

In a study of other subjects having morning off, subjects getting 50 mg inhaled levodopa came on in 25 mins vs 35 on placebo, and an on event was 35% more likely than after placebo.

The C max of the higher strength (50 mg) is 500 ng/ml while 100 mg oral achieves 700-1000 ng/ml. As the authors comment, there are no head to head comparisons of inhaled versus oral.

The authors conclude that inhaled levodopa is an option for rescue therapy.

Journal Club discussion

The data demonstrate only biological efficacy, and it does seem fast acting. There are not data on the quantitative effect versus oral levodopa or even the speed of action versus oral levodopa, dispersible levodopa or subcutaneous apomorphine. There appears to be no carbidopa or benserazide, so patients are going to be needing to take oral levodopa at around the same time or at least have had a decent amount fairly recently.

On a google search in March 2023, the cost is $1223 for 60 capsules of what is stated to be 45 mg but I think this is equivalent to 35 mg in the study. This might seem about a 1 month supply but no – one dose is two capsules. Patients had an average of 2 doses a day and the company specifies a maximum of 5 doses a day or ten capsules which is $204 a day!

In comparison, oral levodopa is $13 for 90 100 mg tablets. The 100 mg appears to be up to double the strength of the higher strength inhaled, which we presume is 2 capsules. So the cost of inhaled levodopa is over 500x greater than standard treatment. It does seem almost provocative not to attempt a comparison with standard care so much cheaper. We can surmise that perhaps inhaled might work about 15 minutes more quickly. The magnitude of benefit is demonstrated biologically but there is no comparison with oral. Measurements were made when the patients were rendered off to maximise the benefit rather than in a real world scenario dealing with unpredictable sudden off. Would the plasma concentrations as little as half that of 100 mg of oral really rescue such patients?

Regarding safety there are one year data. No worsened lung function was demonstrated but this does not mean there is evidence of no side effects unless the study was powered and the type B error quantified. We have many examples of preliminary data not showing the dangers revealed subsequently on post- marketing surveillance.

Posted in Parkinson's Disease, Uncategorized | Tagged , , | Leave a comment

Revolution in acute ischaemic stroke care: a practical guide to mechanical thrombectomy

Posted on March 24, 2023 by dulcetware

Introduction

Stroke is the most common cause of disability in Western Countries, and its lifetime risk is up to 25%. While managing acute stroke patients in hyperacute stroke units is regarded as having benefits for short and long term outcome, specific therapeutic options are limited. The first major option for treatment of ischaemic stroke was intravenous thrombolysis, paralleling its previous development in acute myocardial infarction. However, while use in the latter indication was widespread in the 1990’s, it has only been widely for stroke in the last ten years. This is probably because of the narrower therapeutic window and the more severe consequences of haemorrhagic complications in the brain. In addition, its benefits are actually relatively modest. In randomised clinical trials on use within three hours (bearing in mind that in the first hour a stroke often spontaneously recovers – termed a TIA), using NICE data (June 2007) 15% of patients have an improved outcome, but with a brain haemorrhage risk of 7% greater, and a major haemorrhage rate actually considerably worsening the stroke (>= 4 points on the National Institute of Health Stroke Severity Scale) of 1.7% . When delivered between 3 and 4.5 hours after stroke onset the benefits were not even clearly significant.

So it is not surprising that there has been a move, just like in cardiology a decade or two earlier, away from intravenous thrombolysis and towards direct intra-arterial catheter treatment. This article discusses this new treatment and the ramifications for delivery of such a service.

The paper, Revolution in acute ischaemic stroke care: a practical guide to mechanical thrombectomy, summarises recent evidence in favour of this treatment and the infrastructure required to manage patients in this way.

The Procedure

While the first such devices were approved for use in 2004, technical developments and the improved expertise that comes with experience show in recent studies published since 2010 on new generation devices that they yield major improvements in outcome. The HERMES collaboration meta-analysis revealed that 46% of patients had a good outcome with functional independence (grades 0-2 on the Modified Rankin scale) compared with 26.5% on best medical treatment. Most of the patients in both groups received intravenous thrombolysis, since in most study protocols patients had iv thrombolysis before going on to have thrombectomy an hour or so later. Mortality and the risk of brain haemorrhage did not differ between the two groups. The benefit seemed still to be present in patients over 80, and still present when patients did not receive iv thrombolysis, though the numbers in this case were smaller. While the window for thrombectomy was within 6 hours, there may still be improved outcomes up to 7.3 hours after symptom onset, but in generation faster intervention leads to greater benefit.

The procedure involves a number of variations depending on the Neuroradiologist and the particular nature of the thrombus. It may be done under general anaesthesia or local anaesthesia and sedation with anaesthetic support. A large gauge catheter is directed to the internal carotid via a femoral puncture, and an intermediate catheter inside it is directed to the Circle of Willis. Then a microcatheter inside that serves as a guide wire to the actual clot. The microcatheter is then removed and a stent retriever is placed within the clot, and pulled back to draw the clot to the intermediate catheter. Suction is applied to this catheter to remove the clot entirely. Some techniques involve directly removing the clot by suction on the intermediate catheter. A balloon may be located on the distal end of the clot to prevent forward movement. When removing the clot reveals a tight lumen, there is the further option to perform angioplasty or stenting to open the vessel. The same can apply to a more proximal carotid stenosis occurring in tandem with the more distal thrombus.

Complications

The main complications are technical, including vessel perforation (1.6%), other symptomatic intrcranial haemorrhage (3-9%), subarachnoid haemorrhage (0.6 – 5%), arterial dissection (0.6 to 3.9%), or emboli distally (1-9%). In addition , there can be vasospasm or issues related to the puncture site. While the total incidence is 15%, not always is there any actual clinical adverse consequence.

  • While the time window for thrombectomy is wider than for intravenous treatment, there are other selection criteria that are more strict.
  • There should be a documented anterior circulation large vessel occlusion of the middle cerebral or carotid artery. (There is only limited evidence for efficacy in basilar occlusion.)
  • There should be good collateral cerebral circulation.
  • There should be relatively normal extracranial arterial anatomy from the technical viewpoint of passing the catheter.
  • There should be significant clinical deficit at the time of treatment, but this parallels the criteria for intravenous treatment and a large vessel occlusion with minimal clinical deficit nevertheless incurs a significant risk of clinical deterioration.
  • There should be a lack of extensive early ischaemic change on CT (according to ASPECTS score a threshold of 5). The role of more advanced imaging , eg CT perfusion to establish salvageable brain, is yet to be clarified.
  • Consideration should be given to pre-stroke functional status and the potential of benefit.
  • Patients should have had iv thrombolysis within 4.5 hours of symptom onset.
Posted in Uncategorized | Leave a comment

Testing for COVID-19 Infection

Posted on November 22, 2021 by dulcetware

Accurate testing for SARS-CoV-2 infection, by which we mean testing with few false positives as well as few false negatives, is important not only for clinical management of individual cases but for epidemiological case tracing, limiting spread of infection and informing public health strategies. In the latter situations, tests that are rapid, cheap and easy to perform are particularly desirable.

Two main forms of testing for SARS-CoV-2 infection are in use.

  • Antigen testing, which includes the self administered immunochromatographic lateral flow test (LFT), detects viral coat material and is developed by raising a specific antibody against the antigen target. It therefore measures active production of the viral protein that constitutes the antigen. Developing such a test requires knowledge of how the viral behaves in the host and is reliant upon generating a sensitive and specific antibody to be used for the test.
  • Molecular tests, such as polymerase chain reaction (PCR), loop-mediated isothermal amplification (LAMP) and clustered regularly interspaced short palindromic repeats (CRISPR) tests amplify viral RNA material. These tests are potentially specific for different variant mutations of SARS-CoV-2. The gold standard test is considered to be lab based reverse transcription viral PCR (RTPCR). Rapid testing, such as the LAMP test, skips some of the time-consuming stages of formal PCR and therefore is useful for screening and epidemiology.

Many different studies have reported on the performance of different brands of rapid antigen and molecular test. This article discusses a Cochrane review of diagnostic test accuracy that collected data from 64 studies that investigated 16 different antigen tests and five different rapid molecular tests. In all there were 24087 samples. Of these, 7415 were found on the subsequent gold standard RTPCR test to be positive. No study actually included clinical diagnosis of infection as a standard or criterion of infection.

Antigen tests had a sensitivity in symptomatic cases of 72% (95% CI 63.7 to 79%). The sensitivity was higher in the first week of infection (78.3%) than in the second week (50%). The overall specificity was 99.6% (in both symptomatic and asymptomatic – obviously if symptomatic one wonders what they actually had if we are considering false positives.

Analysing a different way, the positive predictive value (PPV) in comparison was only 84-90% using the best sensitivity brands at 5% population prevalence and at 0.5% prevalence in asymptomatic people the PPV was only 11-28%!

Molecular tests had an average sensitivity of 73% (66.8 to 78.4%) and specificity was 97.2 to 99.7% depending on brand. In most trials the samples were collected in labs rather than in real life conditions. There are no data about symptoms.

For reference, WHO considers >80% sensitivity and >97% specificity an appropriate test as a replacement for a lab test.

The authors note that in low prevalence settings the dramatically reduced PPV value means that confirmatory tests may be required and that data are required for the settings in which tests are intended to be used as well as on user dependence (i.e self testing)

Journal Club Conclusions

Why the huge difference between positive predictive value and specificity?

Specificity = 1 minus false positive rate (FPR)

FPR = false positives/all negative cases (i.e. true negatives and false positives)

So high specificity means false positives is low compared to true negatives

PPV = true positives/all positive results (i.e. true positive and false positives

So high PPV means false positives low compared to true positives

The difference between specificity and PPV is essentially that sensitivity is in relation to the total number of actual cases while PPV is in relation to the total number of positive test results. PPV could be much worse than specificity if the true negatives much more common that true positives. This would happen if infection rates are very low in the population.

Is it disingenuous to quote specificity based on trials where infection rates were 31% when in the real world the infection rates are perhaps two orders of magnitude lower than that? If someone wanted to argue that they were being denied going to work, going to school or going on holiday based on a test where the predictive value was only 11-28%, would they have a good case? Is it worthless as a screening tool? If the policy is to go on to a molecular test if the result is positive, is this also invalid if the PPV of the molecular test is similarly low?

Clearly, the use of tests should be tailored to the information that can be provided. In an outbreak of high prevalence, one wants a sensitive test to pick up people who might have infection after contact tracing. One could have a specific screen and lab test only those negative to make completely sure. The priority is not to miss positive cases.

If medicolegally one wants to prove workers were not source of a case or outbreak, when the prevalence of infection is low, one may as well go straight to lab testing as there are too many false positives in such a situation.

In a situation where someone wants to question a positive result, it is not clear that rapid molecular testing is superior to antigen testing, and a lab based PCR may again be necessary. as specificity not clearly higher than some molecular tests.

There are also biological as well as statistical issues. For example, antigen tests may theoretically have false positives if the nature of the generated antibody to a viral coat antigen is not clear. If the trial was done in the summer time with no winter flu or coronavirus common colds in a setting where one in three subjects have COVID and none has any other type of infection, is the generated antibody really shown to be specific for SARS-CoV-2? The same may apply to demographic factors, such as expecting the test to have the same performance in children or in care homes as in a healthy adult test population.

On the other hand, a molecular test may correlate better with actual infectivity than a bit of residual RNA and therefore be more biologically useful for epidemiological control.

Finally, in extremely high prevalence trial populations where there is no actual clinical corroboration, the absolute reliance upon lab PCR as a gold standard may be of concern.

Posted in Infectious Diseases | Tagged , , , | Leave a comment

Distinguishing Encephalitis from Encephalopathy

Posted on December 11, 2020 by dulcetware

encephalitis

Encephalitis may be defined as infection or inflammation of the brain substance, resulting typically in disturbed sensorium and perhaps seizures or focal neurological deficits and sometimes pyrexia. Encephalopathy in the other hand represents disturbed sensorium not due to an infective or inflammatory cerebral process and its causes range from toxins and drugs to metabolic upset, non cerebral sepsis, cerebral hypoperfusion and post-ictal states.

Distinguishing the two is important because considerable morbidity and mortality is associated with delayed treatment with appropriate antiviral, antibiotic or immune therapy for encephalitis and with delayed treatment of the various causes of other causes of encephalopathy.

The paper presented, “To what extent can clinical characteristics be used to distinguish encephalitis from encephalopathy of other causes? Results from a prospective observational study” by Else Quist-Paulsen et al., attempts to use clinical and rapidly available investigatory findings to distinguish the two conditions by a prospective observational study on 136 patients.

They identified candidate patients on the basis they had a lumbar puncture, and then excluded those with no evidence of encephalopathy. Their criteria for encephalitis were:

  • Pyrexia
  • Encephalopathy > 24 hours with no other cause identified and 2 of:
  • CSF WCC >=5 x 106/l
  • New onset seizures
  • New onset focal neurological findings
  • CT/MRI consistent with encephalitis
  • EEG consistent with encephalitis

The gold standard by which to gauge their test would surely be a definitive diagnosis but, as is commonly the case in clinically suspected encephalitis, such a diagnosis was only made in 10 of 19 patients. In some of the patients with non-encephalitis encephalopathy, the diagnosis was also vague, e.g. “aseptic meningitis” (which could be encephalitis), “epilepsy” (which could be autoimmune encephalitis), “headache/migraine”, “unspecified disorientation or coma”.

Subsequent analysis of specific features in the two groups then becomes somewhat difficult because the criteria themselves become the gold standard and because some specific features were in themselves their criteria. Interestingly, systemic features of infection such as raised blood white cells or CRP, argued against encephalitis because general sepsis was a common cause of encephalopathy. Nausea and personality change were more common in their encephalitis group.

They used ROC curves to look at the predictive value of these specific features and their combinations, but these were again based against their “testing variable”, their criteria, not on some objective gold standard. It would have been better to look at them only in the 10 diagnosed cases rather than all 19, but then the total number of cases would be even lower.

The diversity of diagnosis of their cases was interesting, especially that Lyme disease and TB were as common as VZV and more common than HSV. Only one of their cases had NMDA receptor antibodies, but we do not know that all the patients had this test and a full battery of other autoimmune antibody tests. Many might have been put in the encephalopathy with seizures category. Since encephalitis can be associated with meningism, some “aseptic meningitis” patients might have been viral but with negative testing, or even autoimmune with a migrainous headache and stiff neck.

The group felt that the study was very worthwhile but a more clear guide as to which cases of encephalitis warranted antimicrobial therapy  or immune therapy would be the clear goal. This would require clarity on the gold standard diagnosis and many more patients.

The Journal Club discussion on which this post is based was presented by Dr Aram Aslanyan, Specialist Registrar in Neurology at Queens Hospital, Romford, Essex.

Posted in Infectious Diseases, Inflammatory/ Auto-Immune Diseases | Tagged , | Leave a comment

Making a Differential Diagnosis using a Clinical Database

Posted on November 20, 2020 by dulcetware

Background

A great deal of time is spent in medicine reading and writing case reports. Essentially, clinical features are listed and a diagnosis made. Excluding those cases that point to a novel means of treatment, a case report is often noteworthy simply because the diagnosis is rare, or because the clinical features were most un-likely to be associated with the diagnosis. This hardly seems a reliable method of archiving medical knowledge.

Much less time is spent on attempting a method of diagnosis that is more systematic than the recalling of case reports. One can see that if one did wish to move medical diagnosis into the information age, natural instinct would be to use an internet search engine to enter a list of clinical features and see what disease diagnoses were associated with these terms. Unfortunately, internet search engines concern themselves only with the popularity of search terms and because of the dominance of case reports such practice may be likely to throw up the least likely cause of those features, or that which is most “titillating” to those who most perform internet searches.

There have been attempts to provide a more balanced means of linking clinical features with diseases and hence making clinical diagnoses. Rare disease with a large number of different clinical features are least easily diagnosed by clinical experience or key investigations, and so the focus of these attempts has been on rare genetic diseases using ever-expanding databases such as Orphanet, Online Mendelian inheritance in Man (OMIM) and the London Dysmorphology Database and the Pictures of Standard Syndromes and Undiagnosed Malformations (POSSUM).

One method of searching for clinical features on these databases is simple text matching. A way of quantifying the match is the feature vector method, which calculates the mathematical overlap between the Query (the clinical features of the case) and the Disease (the clinical features of the disease). A vector of the query is calculated with dimensions for each feature and a value of 1 if present and 0 if absent. The same is done for the disease. The dot product of the two vectors is the strength of the match (a 1 for both query and disease will sway the two vectors in a common direction, and a 0 for both will leave their relationship unchanged, while a 0 and a 1 will make one move away from the other).

A potentially better quantification of matching is to take into account the different specificities of different clinical features. If a clinical feature is present in only a few diseases, its annotation (the linkage of a clinical feature to a disease) is more specific for that disease (in database terms this is called the information content (IC)) and so that linkage should have more weighting. The IC is simply the negative log of the frequency of the annotation. For example, AV block is a term that annotates 3 diseases in the 4813-disease OMIM database. The frequency is 3/4813. Loge of this is -7.38 and minus loge of this is 7.38. A much more general term will have many annotations and a much lower negative log, tending towards zero. The ICs of all the clinical features of the query can be summed or otherwise combined to provide an overall match.

The authors of the presented paper have described a further refinement of this method. This is called the Ontology Similarity Search (OSS). Instead of simply matching the text of terms, they fit clinical features into a standardised language within an ontological framework. This means that the features are related to one another in a hierarchy, with more general terms higher in the hierarchy and more specific subcategories of those general terms lower in the hierarchy. While “parent” terms obviously have many “child” terms, child terms can also belong to multiple parent terms. For example, optic atrophy could be a child of demyelinating disease and also a child of visual disturbance. Their ontology is called the Human Phenotype Ontology (HPO) and has around 9000 terms.

The advantage of using the ontology is that if a clinical feature of a case does not fit the clinical features of the disease, but shares a parent term with one of the features of the disease, instead of scoring a zero match, this scores as a match but less so than if the match was with the specific terms. The method specifically find the most informative common ancestor of the two different clinical features, and uses the IC of that term. Being a more general term, it will be a feature of more diseases and so have a lower IC. (In the database, ancestor terms are implicitly annotated when child terms are annotated.) The overall strength of match is the average of all the ICs – there will always be a IC for each feature, even if it is just that they are both a feature of “any disease”, which of course has an IC of zero and would bring down the average.

Summary of the Paper

The presented paper, Clinical Diagnostics in Human Genetics with Semantic Similarity Searches in Ontologies by Köhler et al. (Am J Hum Genet. 2009 Oct 9; 85(4): 457–464), describes a further refinement of the method using a statistical treatment. For a given disease, if random clinical features from the HPO were selected one would expect a lower OSS score than for a patient who actually had the disease. If the OSS for random features were repeated many times, a distribution would be created and so one could then look at the real patient OSS and determine a p-value on this distribution. If the real OSS was higher than 95% of the random OSS scores, the p-value would be lower than 0.05 and indicate a likely match. Furthermore, if the same features were compared with different diseases and their random OSS distributions, a ranking of the likelihood of diseases could be determined by ranking the corresponding p-values. They call this the OSS –PV.

Since they considered it too onerous to enter, within the framework of the terms of the HPO, the clinical features of real patients with known diseases, they used simulated patients. This was done for 44 diseases, where they created a “patient” having a disease with a selection of the clinical features of the disease weighted by how commonly those features were found in that disease. For each disease 100 patients were created, so if from the clinical literature a feature is found in 1% of cases with the disease, 1 of the 100 simulated patients would have that feature.

They added “noise” to the process by adding to the patients some random features that were not part of the disease, and “imprecision” to the process by replacing some features with their HPO parent terms.

Then they looked at the rank position of the true disease among all the 5000 or so database diseases found by the different methods. The closer the rank position to the true position (first!), the better the method performed.

Unsurprisingly, the performance of the feature vector method, as shown by box plots of rankings for all 44 diseases tested, was found to suffer when imprecise terms were used, because that was the point of using the ontological system. The OSS-PV method more modestly outperformed the raw OSS method when noise and imprecision were added.

As the authors point out, the OSS method potentially suffers from the fact that it only matches query terms with disease terms. If a disease also had many terms that did not match the query terms, surely the overall match would be less specific. This can be taken into account by performing a symmetrical similarity search, where the OSS is the average of the matches of the query to the disease and the matches of the disease to the query. However, they did not use this method in their presented data, only stating that when they used it the symmetrical OSS-PV still significantly outperformed the feature-vector method. They do not state that it still outperforms the symmetrical raw OSS.

Another point raised by the paper is that if one finds on a disease search that no disease fits the features with a p-value less than 0.05, exploration could be made of other clinical features, or child features of the entered clinical features that would have a higher information content and provide a more significant match. Going back and looking for a specific feature, or performing a specific investigation, would be an example of this.

Journal Discussion

As described in the introduction, any attempt to quantify and rationalise differential diagnosis should be lauded and this paper clearly describes progressive refinements of this process. It is almost negligent to have all the data available on thousands of diseases and not to use them because the unaided human mind simply cannot store so much information.

However, a number of further refinements and limitations present themselves.

First, the matching of terms is still semantic rather that systematic. While a knowledge-based approach, it nevertheless does not rely on understanding of disease pathophysiologies and pathognomonic features. Some clinical features that share a close parent may in fact best distinguish diseases rather than be considered loosely positively associated features. This may apply particularly in neurology where there is a more systematic approach. For example, upper motoneurone lesion and lower motoneurone lesion may be considered together and share a common parent in “motor neurone lesion”, but apart from the case of motoneurone disease, they split the differential diagnosis more than upper motoneurone lesion and no motor lesion at all. They are semantically similar but nosologically opposite. Horizontal supranuclear gaze palsy and vertical supranuclear gaze palsy may share a strong information content parent, but may be the feature that best separates Gaucher disease from Nieman Pick disease.

This leads to the second point. The frequency, or sensitivity, of a clinical feature in a disease is not considered, although ironically considered when creating the simulated patients with the 44 tested diseases. In large part this reflects the lack of clinical data in the databases themselves. It is regrettable that case reports are not combined into case series which contain information on the frequencies of occurrence of clinical features, or when there are case series, these data are not actually collected systematically. If a clinical feature occurs in 1% of cases of one disease and 100% of cases of another disease, clearly the annotation of the feature for the second disease should be considered far stronger than for the first. Instead, because there are no such data, they are given equal weight; the weighting only considers whether or not the feature is also found in a number of other diseases, not how commonly it is found in those diseases.

There is no consideration of how common the disease is in the first place. While restricting themselves to rare and genetic diseases by definition, there can be a frustrating tendency for searches to throw up the least likely diagnosis. It is often the case in practice that the clinician does not know in advance that the patient has a rare genetic disease, and a diagnostic tool should be most useful to those with least intimate knowledge of the database. Thus, when entering the features dystonia, spastic hemiparesis and spastic dysarthria in a case of cerebral palsy, it comes as a surprise when the top diagnosis is cleft palate-lateral synechia syndrome.

Finally, the methods assume that clinical features are independent. In fact, many clinical features are strongly interdependent; they especially occur together. The association of the second feature is not really very additionally informative if the first is present. This problem would be common to most forms of differential diagnosis calculators, including those using Baysian methods, and could only be solved if there were data on the interdependence of clinical features in different diseases; currently it is hard to find even raw frequency data for most diseases.

The point that the authors raise about using their App to find features that would be more specific in making a diagnosis is an interesting one, and opens a new approach to diagnosis and refinement of the process of often expensive and sometimes risk-associated investigation. One could imagine the improvements in medical care that would arise from use of an App that gave a differential diagnosis based on initial clinical information and then showed the relative power of different investigations in narrowing that differential.

A further use of these methods would be in creating diagnostic criteria. While clinical practice is rightly focused on the most likely diagnosis in a patient, clinical research is focused on a group of patients where the diagnosis is certain, i.e. specificity at the expense of sensitivity. Currently, diagnostic criteria seem to be set largely by “workshops” – gatherings of the great and the good usually in an exotic location who draw up a list of features, create two categories of importance and then decide how many features are required for a “definite diagnosis”. Using a quantified method such as that described in this paper for every study patient and including only patients where the diagnosis reaches a threshold p-value score would seem to be a far more reliable method.

The paper on which this journal club article is based was presented by Dr John McAuley, Consultant and Honorary Senior Lecturer in Neurology at Queens Hospital, Romford.

Posted in Genetics | Tagged , , , | Leave a comment