New Zealand: Dr Matthew Phillips, neurologist

Hutan said:
The FNDA website appropriates Māori culture:
We choose the Kaku Kiwi, our nation's ceremonial kiwi feather cloak, to represent our mission.
Click to expand...
Click to expand...
... and manages to mess up the spelling in the process. Should be kahu not kaku. Though perhaps fittingly kaku means (as per Te Aka dictionary):
1. (noun) Pieces stripped off in the process of dressing flax.
2. (noun) Dried leaves or other vegetable refuse, such as that often deposited by a flood.
3. (noun) A rough cape made of such pieces.

Rough indeed
 
Subject: Clarifying my views

Hello everyone, I’ve become aware of this discussion about comments I made in the Goodfellow Unit podcast. I appreciate that many of you have strong experiences with ME/CFS and concerns about diagnostic practices. I deeply respect the experiences of people living with FND and other complex conditions like ME/CFS, and I advocate for a better biomedical understanding of these conditions, which is why I started the FND clinic at Waikato Hospital in the first place.

In my talks I often present FND as one of the oldest, most misunderstood, and challenging areas in neurology. I think it is often disrespected as a valid diagnosis by both doctors and patients, which is unfortunate. To be clear - I do not see these conditions as psychiatric or psychological. They are very complex, not easy to diagnose, and we still have much to learn. Moreover, FND should never be used as a diagnosis of exclusion or a way to dismiss patients. A thorough workup to identify potential co-existing structural or metabolic problems is essential. Many of the most complex patients have both FND features and underlying structural (neuropathy, disc protrusion etc), metabolic, or other organic issues - missing the latter, if present, is a serious mistake and negatively impacts outcomes.

In our FND clinic at Waikato Hospital, a number of cases are not “pure” FND. These more difficult cases have co-existing pathology that has been downplayed or missed by previous doctors. We directly address this with thorough assessment and individualized management. Many patients who arrive very frustrated and skeptical report meaningful therapeutic benefit and improved function afterwards. I devote 1-2 hours per patient, so as to give enough time to do my best to get it right.

While it is important not to miss a structural or metabolic problem, I have also seen the opposite problem - colleagues who did not use a good history and examination technique rely on multiple investigations (even whole-body scans) to try and find a pathology that does not exist. This too is intensely frustrating for patients, as they simply feel like they are a guinea pig, getting lots of tests over months or years but no clear diagnosis - which many people need in order to get a treatment plan together and improve.

I also think it is unfortunate that many people view FND as a psychological or “made-up” problem, when the evidence points towards a distortion of predictive processing - a concept with strong neuroscientific evidence. This distortion can occur in isolation (no apparent trigger), or it can be triggered (by an infection, medication, or something else), or it can occur in conjunction with another structural or metabolic issue (sometimes called functional overlay). If the latter, it can be difficult to assess how much the distorted predictive processing element contributes to the measurable pathology and vice versa - the direction of causality likely varies from one person to the next.

Regarding the “list sign” I mentioned in the podcast - it can be a useful clinical observation as I mentioned, but it is only one sign among many. I regularly teach students and trainees that it should never be used in isolation to diagnose FND. So if my comments came across as overstating its importance, that was not the intention.

When I refer to ME/CFS, fibromyalgia, IBS etc. as “cousins” of FND, I simply mean that they frequently overlap and may share elements of abnormal brain processing (predictive perception of internal signals). Both structural and metabolic factors can play important roles, sometimes triggering subconscious brain processing changes, but also, there is emerging evidence that subconscious brain processing distortions can likely trigger metabolic, immunological, and other pathology in some cases. Again, the direction of causality likely varies between individuals. This needs a lot more research.

On the research side, I’m focused on exploring metabolic therapies (ketogenic diets, fasting protocols, etc.) as potential adjuncts for difficult neurological conditions such as Alzheimer’s, Parkinson’s, and glioblastoma. The early metabolic trials are promising - especially given how tough these disorders are to treat - but I fully agree they need larger confirmatory trials and I usually stress that when I discuss our work.

Helping people with FND and other tough disorders can be very rewarding when done carefully and individually. While we have a fair way to go before we can say we understand them well, the therapeutic approaches I mention do help many - not all of course, and I am always up-front about that. Still, my main goal is to help improve neurological outcomes for patients, any way I can.

Anyhow, thank you for your time and the feedback. I am happy to answer specific, respectful questions about my work or views (email is best, which can be seen on any of my publications).



Sincere regards,



Matthew Phillips

Neurologist, Waikato Hospital
 
Dear Matthew,

I am an emeritus professor of connective tissue medicine, as you probably know. (I introduced B cell depletion for autoimmune diseases like MS). I have been studying ME/CFS for ten years now, largely through scientific discussion with people with the disease but also as part of research collaborations.

The problem for members here is that the explanation you give does not add up.

We don't know why these conditions are referred to as 'complex' since that term has no real meaning if dealing with a process nobody understands, other than of course a general euphemistic meaning of 'psychological overlay'. There is way too much doublespeak in this area.

We do not understand the physiology of ME/CFS. The genetics has pulled out CA10, OLFM4, BTN2A1, which give us clues, but there is no story beyond that.

In relation to FND, the problem members here see is that almost all medical colleagues publishing on it clearly hold two simultaneous and rather contradictory views. One is that it just means a problem in signal sending patterns without structural change. The other is that it is likely to be based on childhood trauma or other psychological input and is best managed psychologically. You may not hold the latter view but many of those proposing a predictive coding model explicitly do.

As a physician who trained in neurology I find it hard to see the advantage of the FND category as a way of bundling together the many cases we see without identifiable structural pathology who almost certainly have a wide range of completely different problems.

Then there is the predictive coding idea. You say there is 'strong scientific evidence' for this. Sorry, but the level of scientific debate here is higher than that. We have known about prediction in perception and action since Pat Merton made the black and white films of manipulating his own eye movements that he showed me as a neuroscience student 50 years ago, together with his work on gamma efferents. It was clinically obvious to physicians long before, that prediction was involved in some aspects of cognition - but there are lots where it isn't. It is now an empty buzzword rather than physiology. So, yes, there is strong scientific evidence for predictive processing, but none whatever that it is the basis of 'FND'. In fact you will find a discussion thread here where the agreed conclusion is that predictive coding gives the wrong prediction as an explanation for FND. It predicts the opposite of the clinical problem. At a more basic level we have often discussed here that you cannot explain pathology with normality. you have to identify the deviation from normality that defines the abnormal process. Without doing that you have no useful explanation.

I would be interested in your further thoughts. As you can see from Trish's question, the patient members here are sharp. The level of critical analysis is way above what we get in a medical environment. There are members here responsible for setting and running up major science projects involving neurobiology. If there is strong evidence people want chapter and verse.

All best wishes

Jo E
 
Kia Ora @Matthew Phillips.

My questions relate to working in a challenging clinical field that is also seeing fairly rapid changes in the understanding of its fundamentals, and how you navigate this.

For example, previously it was held that there was no structural abnormality underlying FND. That it was purely functional, a disorder of processing. This was often explained to patients using the analogy of "it's software, not hardware". FND researchers have been building the case over the last five years or so that there is a structural abnormality, demonstrable with neuroimaging, and that possibly this forms the substrate on which the disorder is derived (vs being a secondary effect). Potentially you have clinic patients that are being cared for through this transition and into the future. How do you try to navigate the clinical relationship when things are in such flux, potentially even to the point of reversing the original premise? (A bit like "low fat diet everyone!" to "you know guys, it might be the sugar that's bad.")

Related, you mentioned the comorbid conditions, and canonical examples might be Parkinson's disease or MS with comorbid FND, or "with functional overlay". In the case of MS, research is now indicating that the normal appearing white matter may actually be abnormal, and that in fact the NAWM is the part that's actually driving the main symptoms of MS far more than discrete lesions or regions that might be visible on MR. That sort of sea change in clinical understanding, new even in the last 12 months must be challenging.

Do you envisage real clinical utility in more advanced neuroimaging translating from early research to practice? If FND is such a major component of neurology presentations, that's unlikely to be realistically achievable with the MR fleet available in most countries. How would you go about choosing who would benefit from such a confirmatory test (assuming the research trajectory continues as suggested)?
 
Last edited:
Thank you for your post @Matthew Phillips and for joining us here.

Matthew Phillips said:
I devote 1-2 hours per patient, so as to give enough time to do my best to get it right.
Click to expand...
That's great. However, your advice to GPs from one of your webinars mentioned above suggests that they start thinking about a diagnosis of FND if the patient is still mentioning symptoms after 5 minutes of talking to the GP. Because, you say, those 'consult times are so damned short'.
​
"In terms of what to look for, well I would say as a GP —well the consult times are so darned short — if you spend five minutes talking to a patient and they're giving out symptoms and they're still going and you're very confused then chances are creeping up that it's one of these. If they have a list of symptoms — that's known as the 'list sign' — that's actually an extremely reliable indicator that you're probably dealing with a functional neurological disorder. Or one of its cousins: there's other brain processing disorders from my perspective as well, so like chronic fatigue syndrome I think falls into this category, sort of a cousin of FND. As does fibromyalgia and irritable bowel syndrome."​
​
“Most doctors, once you get to three or four symptoms, our brain starts to turn off and go arrrghh I can’t localise it. There’s just too much stuff going on here, why does this person have, you know, the list”​

It seems, from what you say there, that the diagnosis of FND is much more about the doctor's confusion and overwhelm than anything else.

I think it's useful to put yourself in the position of the patient with a health condition that doctors can't immediately diagnose. The patient is likely to be trying hard to give the doctor all the information that might provide the key that allows a diagnosis, including every symptom they can remember, even though they know that probably not all of them are related to the health condition.

Do you understand that a diagnosis of FND, even if casually given by a GP, brings stigma that will probably negatively impact on the patient's health care for decades? Do you stand by the suggestion that the list sign is an extremely reliable indicator of FND and cousin disorders such as chronic fatigue syndrome and that GPs should consider FND if they are confused and the discussion of symptoms with a patient takes more than 5 minutes? If not, would you consider removing that webinar from the training materials for GPs?



Matthew Phillips said:
In our FND clinic at Waikato Hospital, a number of cases are not “pure” FND. These more difficult cases have co-existing pathology that has been downplayed or missed by previous doctors. We directly address this with thorough assessment and individualized management. Many patients who arrive very frustrated and skeptical report meaningful therapeutic benefit and improved function afterwards.
Click to expand...
How do you know the symptoms a patients has aren't caused wholly by missed diagnoses? What percentage of the patients that come to your clinic do you conclude do not have FND? Do all the patients with a missed diagnosis also have an FND overlay?



Matthew Phillips said:
I also think it is unfortunate that many people view FND as a psychological or “made-up” problem, when the evidence points towards a distortion of predictive processing
Click to expand...
A number of key proponents of FND are very upfront that FND is another name for conversion disorder, that is, definitely a psychological problem. We can find the quotes if you like. Here is some detail about coding, also see here, here and here.

Dx Revision Watch said:
The term the American Psychiatric Association (APA) chose for use in DSM-5 is "Conversion Disorder (Functional Neurological Symptom Disorder)" and the disorder category was moved under the Somatic Symptom and Related Disorders category block:​
​
Somatic Symptom and Related Disorders​
​
Somatic Symptom Disorder​
Illness Anxiety Disorder​
Conversion Disorder (Functional Neurological Symptom Disorder)​
Psychological Factors Affecting Other Medical Conditions​
Factitious Disorder​
Other Specified Somatic Symptom and Related Disorder​
Unspecified Somatic Symptom and Related Disorder​
Click to expand...

Do you agree that conversion disorder is a synonym for Functional Neurological Disorder? If not, could it be that your idea of FND is different to other people's?


I think it would help us understand what you mean by FND if you could explain your approach to treatment.

Thanks again for your presence on the forum.
 
Last edited:
Some more questions and comments for you, but the last of mine for a while. I don't want you to feel overwhelmed with posts to respond to, please take your time.

Matthew Phillips said:
On the research side, I’m focused on exploring metabolic therapies (ketogenic diets, fasting protocols, etc.) as potential adjuncts for difficult neurological conditions such as Alzheimer’s, Parkinson’s, and glioblastoma. The early metabolic trials are promising - especially given how tough these disorders are to treat - but I fully agree they need larger confirmatory trials and I usually stress that when I discuss our work.
Click to expand...
I think it would be useful if you could address the concerns I expressed about your trial of a ketogenic diet up thread - see here.

Matthew Phillips said:
Still, my main goal is to help improve neurological outcomes for patients, any way I can.
Click to expand...
You sound as though you genuinely care, which is why I want you to know what your interventions and your labels actually mean for patients.

I don't know if you yourself have tried a strict ketogenic diet for any length of time? If you have, you will know that it is fairly difficult at the best of times. Dealing with Alzheimers is definitely not the best of times, not for the patient nor for their probably elderly carer. New foods and different routines are hard to cope with. In order to accept that level of difficulty, there needs to be a clear benefit. That trial of yours simply did not demonstrate sufficient benefit. Specifically, the blinded assessment of cognition did not find a benefit, the subjectively assessed benefits for the completers were extremely marginal.

I would like to understand how you can have done that trial and say that the ketogenic diet is a promising treatment for Alzheimers. I look at the trial report and think about the difficulty imposed on the participants and their carers and the lack of credible benefit and see a trial that suggests the intervention is not worth pursuing further. Every research dollar that now gets allocated to ketogenic diet trials for Alzheimers is almost certainly a dollar not spent on things that might actually help people with Alzheimers.

Similarly, I am not aware of any sound trial that has found benefits from treatments of FND. FND trials also suffer from a range of methodological flaws usually including subjective outcomes in unblinded trials.

You are putting labels on people that stigmatise them - most clinicians see FND and think the equivalent 'hysteria' or 'conversion disorder'. When people with an FND label go into the Emergency Department with a heart attack or kidney stones, that label may well result in them getting less respectful and less helpful care, perhaps resulting in serious harm. The impact on young people who are told that they can't trust that their experience of the world is right is particularly consequential. So, there are substantial downsides to having an FND diagnosis.

Therefore, to justify the diagnosis, there has to be a major upside for the patient - most compellingly, a cure or at least a worthwhile improvement in function. We keep track of the FND literature fairly well here. I don't think we have seen any credible evidence of such an upside. I hope that you will tell us about the studies that convinced you that the treatment you provide is useful.

Perhaps you are relying mainly on your clinical experience when concluding that you can usefully treat people based on an FND paradigm? Perhaps you feel that you are in fact delivering that necessary major upside for your patients? Perhaps you feel that your interventions are so individual, so tailor-made for the patient, that they can't be studied in a proper trial?

It is incredibly easy for clinicians to convince themselves that the treatments they offer work. The history of medicine is littered with interventions that clinicians promoted that have since been proven to be ineffective. And, the field you are working in, with desperate patients who are often grateful to anyone who takes their difficulties seriously, is fertile ground for clinicians drinking their own Kool Aid.
 
Hutan said:
It is incredibly easy for clinicians to convince themselves that the treatments they offer work. The history of medicine is littered with interventions that clinicians promoted that have since been proven to be ineffective. And, the field you are working in, with desperate patients who are often grateful to anyone who takes their difficulties seriously, is fertile ground for clinicians drinking their own Kool Aid.
Click to expand...
Even more than that. Moniz got the Nobel Prize for inventing lobotomies. I don’t know how many «lesser» prizes that have been given to clinicians and researchers that have no idea what they are doing, or that are actively causing harm, including OBEs in the UK.
 
Matthew Phillips said:
the evidence points towards a distortion of predictive processing
Click to expand...
We have been discussing the research and evidence in the literature on this issue for several years, and have not come across any such evidence. More importantly, as was pointed out above, this evidence against points against this model, as none of us expect any of the symptoms and loss of function that results from doing normal things. Literally no one has ever expected post-exertional malaise, the main feature of ME/CFS, and in fact our healthy bodies mostly have reliable evidence that moving feels great.

Psychosomatic ideology has completely dominated everything relating to us for decades. It's very difficult to understand that speculative concepts can ever be applied in real life, with real people, especially so in health care, as no one else does that, no other profession relies on flimsy evidence to this degree. We have seen similar things with assertions of evidence relating to us, how it's "illness beliefs" (even though it's similar to predictive coding where our beliefs actually align against the model), deconditioning (which has never been demonstrated, and in fact in some cases psychobehavioral leaders have actually excused the failure of exercise trials on the basis that the participants, which they recruited themselves, were actually plenty active after all), and various speculative assertions about personality traits, "illness as an identity", and so on.

Basically everything in this ideology is assertive speculation, most of it has been debunked, and it still rules our lives, for reasons that no one can actually coherently articulate. It's all asserted (again this theme of simply being assertive) that this is largely built on evidence-based medicine, and yet there is no such evidence out of trials, which keep failing to demonstrate any real benefits in trials, most of which are very low quality and poorly controlled. It's genuinely one of the most bizarre things happening out there.

This assertive speculation model has been around for decades, and has made so many false assertions, and clearly nothing has changed, everything new is old speculative assertions with tiny variations. Why do you think the current speculative assertions are any better than this? Seeing as there is no firm evidence, nothing out of clinical trials, no clinics that can actually boast of producing meaningfully better outcomes that doesn't require a very strong bias towards accepting those assertions at face value?
 
Hutan said:
Matthew Phillips is Director of Neurology as Waikato Hospital.

He's published on predictive processing (brain function).

He also has a strong interest in ketogenic diets as treatments for neurodegenerative diseases. There's a 2021 paper here with Philllips as first author on a trial of a ketogenic diet in people with Alzheimers Disease.

Note that it is claimed that the study is assessor-blinded. Participants were required to have a






So, three primary outcomes.


1. Addenbrookes Cognitive Examination
This was probably given to the patient by a blinded assessor. No significant differences between the ketogenic diet and the normal diet



2. ADCS- Activities of Daily Living (ADCS-ADL)
This was administered to the trial partner. That is, the person living with then AD patient, the person who was required to '(at least partly) partake in a ketogenic diet' and therefore almost certainly was involved in the preparation of the food and giving it to the AD patient, was the one reporting on changes. It is almost irrelevant whether the nurse or whoever was writing down the partner's responses was blinded to the diet when the partner clearly was not. This was not an assessor-blinded study, despite the claims that it was.

Even with that bias, the report benefit was minimal.

Actually, Table 2 shows us that the participants were assessed as increasing their ADCS-ADL by 0.13 of a point while on the ketogenic diet, it's just that the participants were assessed as decreasing their ADCS-ADL by 3 points on their usual diet. The scale is from 0 to 78. Even a change of 3.13 points on a 78 point scale is deeply within the realms of a placebo effect. Honestly, I'm surprised it was not substantially more of a benefit reported given the obvious motivation of the patients and their partners who completed the trial and the hope they would have had that they could do something to improve things.


3. Quality of Life in Alzheimers Disease


So, this was also assessed by the partner of the patient and so was also the opposite of a blinded assessment. A difference in the change between the two treatments of 3 points on a scale of 52 is also barely noticeable and well within the change we would expect with an unblinded treatment and hopeful participants.



It's pretty clear that the study found no benefit likely to be real. And these are results from the 81% of the participants who started the trial. One participant stopped due to diarrhoea (this being blamed on the patient eating too much coconut oil) and four participants were withdrawn because they refused to alter their diet, causing conflict with their partner.
This is what was said about the participant who ate too much coconut oil:

The 'enthusiastic' trial partner does not sound to have been in a state of equipoise.


This is what Phillips et al reported in the abstract:

And this is what was reported in the Discussion:



Matthew Phillips Researchgate entry says


If we think about the people with Alzheimers disease and their carers, dealing with a difficult new diet is not an easy thing. It can be expensive and time consuming at a time when the carer often has to give up work and when things most need to be easy. The text of Phillips' trial notes that the diet caused conflict between some patient and their carers, and diarrhoea in a person with Alzheimers disease probably was very hard for both the patient and the carer.

This is the advice that Phillips is giving medical professionals via the Goodfellow Unit, claiming as he does so that he is a metabolic neurologist:


I think this messaging is a failure of Phillips and the Goodfellow Unit. Phillips' trial is not strong enough to warrant recommendations that will make life harder for AD patients and their carers.

Given the essentially null results from the trial, the positive spin put on them in the report, the misleading characterisation of the trials as involving blinded assessors and Phillips' ongoing commitment to the idea and willingness to promote it, it looks to me as though Phillips is a man who sees what he wants to see when it comes to ketogenic diets. He also seems extremely sure of his opinions about their utility, even when the data does not support them.

Those two qualities may be relevant to his promotion of the concept of functional neurological disorders and his ability to carefully diagnose the people who come to him for help.
Click to expand...
I find this extremely disappointing and discouraging. If we can't trust peer-reviewed publications in high-quality journals (not sure if this applies here, but presumably?), then how is the general public supposed to access objective information?
 
You must log in or register to reply here.
Back
Top Bottom