Revisiting the 2015 MDS clinical diagnostic criteria for Parkinson's disease

And I have two of the authors today with us. We have Susan Fox from the University of Toronto Western Hospital, and Claudia Trenkwalder from the Paracelsus-Elena Hospital in Kassel and the University Medical Center in Goettingen, Germany. Thank you very much Susan and Claudia for joining us today.

Dr. Susan Fox: Happy to be here.[00:01:00] 

Prof. Claudia Trenkwalder: Yeah.

Dr. Eduardo de Pablo-Fernández: Before we start getting a bit more in detail about the paper, let's talk about the clinical diagnostic criteria published from the MDS. So believe it or not, it's over a decade since they were published and they are anchoring the expert clinical diagnosis. And for this, they require the demonstration of Parkinsonian syndrome on the examination, and then a different combination of exclusion criteria, red flags, and supportive features. And depending on the combination there is two categories defined clinically established and clinically probable Parkinson's disease. Where did this idea about revisiting that diagnostic criteria come from? How do you think this could be improved?

Or what are the shortcomings of the current diagnostic criteria?

Prof. Claudia Trenkwalder: So maybe I can start with this background and history and it was initiated during my presidency at MDS that we saw these criteria [00:02:00] are already out for several years and are used and, but others still use the UK Brain Bank criteria. And is there something that we can improve of this criteria maybe. And they're, I think, very good, very detailed. But sometimes it's time consuming, also for neurologists and especially for clinical trialists, if you look at the many supportive criteria, absolute exclusion criteria, red flags, and as you just said, you have to balance them. Although it's not exactly written how you balance it. There are many combinations that can happen, and so we thought with the new and increasing knowledge of also biomarkers, genetics, which is not included in this criteria, maybe one could improve it. But the [00:03:00] first step of all this is always to look at them and to compare them with data. And to go a little bit back up to Jean-Martin Charcot, Charcot was really the first one who looked at neuropathology and clinical symptoms and compared them already in the 19th century.

And he was also the one who rejected the term paralysis agitans for Parkinson's disease because patients were not necessarily paralyzed or did not have to tremor if they had no tremor, and he redefined the disease and advocated for the name maladie de parkinson, which means just Parkinson's disease. So I think he was the first one.

He [00:04:00] looked at many Parkinson patient at the Salpêtrière, and he looked also at the neuropathology, and then he defined the disease and he defined the symptoms we still using at the core symptoms, rigidity, bradykinesia, and possibly tremor. And he differentiated also the tremor from the multiple sclerosis tremor.

So I think it was very helpful to look at the neuropathology and the clinical symptoms and as the MDS criteria there established using consensus of experts and literature review, but not neuropathology. We thought the first step might be to look at these many items, and these are more than six items from the supportive about 10 and 11 [00:05:00] exclusion and red flags. How valid are these items? Can we find data to prove them? Or maybe to reject them or modify them. And therefore these idea came and we discussed this in MDS and finally was agreed to start an ad hoc committee of MDS criteria for Parkinson's disease. And the goal of the project was to develop an initial statement to provide a critique of the current criteria.

Outline the needs of the field, maybe also including biology, genetics in the next step, and to suggest how a path forward could look like. And I think this was the background of the paper and then the [00:06:00] work started with the neuropathology and the MDS committees always suited our balance for gender and geography, and I think we had a good mix and a real expert committee and therefore we could do the work quite, I think, an efficient atmosphere.

Dr. Eduardo de Pablo-Fernández: And you have done so far, like a huge amount of work. So for this article you reviewed all the available pathological studies, clinical pathological studies with a confirmed diagnosis of Parkinson's, including also patients with atypical Parkinsonism. In this incredible amount of data, you try to apply the MDS criteria with the available clinical information. So how did you find that with the retrospective application of the criteria, was that challenging? What were the challenges of this methods and how does affect your [00:07:00] interpretation of the results?

If you can expand a bit about the methods.

Dr. Susan Fox: Yes. Maybe I could, take that one on. Yes. I think what we were really trying to do was look at each of the individual criteria as you've mentioned, the supportive criteria, the absolute exclusion criteria, and the red flags individually. And just to confirm whether these held up when we had the pathological confirmation, we do know that prior studies, including studies that you've obviously been authored to have looked at the overall criteria. But what we were really trying to extract out was the individual criteria and how useful these were. And I think that's where we did run into some difficulties, challenges in actually applying the strict wording of the criteria with the papers that we extracted.

And I think some of the key challenges were some of the very specific wording. For example, the timeframe that was very specific [00:08:00] within the red flag. So for example rapid progression of gait impairment to wheelchair within five years of disease onset, severe orthostatic hypertension within the first five years.

And there were a number of these where there was a specific timeframe and often that wasn't captured within the clinical data. So that made it somewhat difficult to be very specific. The other difficulty was within the supportive criteria, I would say where the definition says a clear and dramatic beneficial response to dopaminergic therapy.

So what do we mean by clear and dramatic? I think there are difficulties here. I think as clinicians we know what we mean, but when we're trying to apply this in a more rigorous setting, it actually becomes slightly challenging. Similarly with motor fluctuations, unequivocal and marked ON/OFF predictable end of dose wearing off.

So this often wasn't necessarily clearly defined, so we had to make the best of what we could [00:09:00] within the clinical definitions. So these were some of the, I think, the main challenges that we faced in terms of the methodology that is applying the criteria to retrospective case series to try and determine how valid these were in the pathologically confirmed cases.

Dr. Eduardo de Pablo-Fernández: Excellent. And overall the individual items showed like they hold to this validation. They are present with more significantly in the Parkinson's group compared to other alternative diagnosis. But if you have to summarize the results, what items of the MDS criteria could potentially be improved? If you can give us a summary of the main conclusions.

Dr. Susan Fox: Yes. So I think, again, taking the individual items, which was our aim here. So if we think about the supportive criteria, the ones that sort of held up best was levodopa response. Within the caveats that I just mentioned. [00:10:00] Anosmia or hyposmia was also a very specific symptom, although there wasn't too much data on the atypical Parkinson's.

Things that could be taken out within the 2015 criteria was the MIBG-spect, where in fact we found no studies that reported any clinical pathological correlation with this investigation. So it may be perhaps, a redundant, we can talk about further biomarkers later on, but obviously that is an initial biomarker that was included in the 2015 criteria.

Looking at the absolute exclusion criteria I think some of them can be useful. You're not likely to have a patient with Parkinson's disease sat in front of you if they have cerebellar signs, absence of observable response to a high dose of levodopa, and have a normal DaT-SPECT presynaptic imaging.

So those would all exclude Parkinson's, and I think those were probably useful absolute exclusion [00:11:00] criteria. The one within that category that we found perhaps more challenging was the description of behavioral variant, frontotemporal dementia frontal cognitive changes. This was, again, a very non-specific symptom that has been described in a number of series, but rather vague and perhaps needs better definition.

The other one where no one actually had ever really reported in clinical series was Parkinsonian features restricted to only in the lower limbs for more than three years. Perhaps it's a very specific symptom people are looking for and perhaps wasn't necessarily documented in any of the series that we pulled out.

And then finally, within the red flags, I think some of the useful red flags to think about. Another diagnosis, not Parkinson's disease were, inspiratory strider, which clearly was seen in MSA. Genitourinary urinary issues, more common in MSA PSP, and then early bulbar symptoms, again, more common in PSP [00:12:00] MSA.

So those three seem to be quite useful. The one I think that's, probably because it's the vaguest, and we found no studies that had looked at this was the absence of common other non-motor symptoms despite five years of disease duration. We know how common non-motor symptoms are in Parkinson's disease and probably also within the atypical Parkinson's, but that particular subcategory there was no real clinical information. So perhaps it needs either to be better defined or perhaps just even taken out if it's not gonna be useful within the MDS criteria.

Dr. Eduardo de Pablo-Fernández: So I take from this that then some items can be, could be potentially removed or simplified and make the diagnostic criteria a bit more easy to apply in clinical practice. But I guess it's challenging to, as you said, to be a specific and in the definition so that we can, the term is some clinical criteria with precision and it's difficult to keep that balance.

Also when you analyze the [00:13:00] items individually, there is only like a few that can be completely diagnostic or can rule out or ruling Parkinson's disease. So I guess it is, very challenging to optimize the MDS criteria. Some people may argue that we may have reached the ceiling of a diagnostic accuracy with clinical methods.

I dunno how you are planning to take this forward, or how do you think that this could be implemented in practice, how you're planning to make those changes into the MDS criteria in a practical way that can be easily used in clinical practice.

Prof. Claudia Trenkwalder: Maybe I can add something to this as we did not talk yet about co pathology. And it was very useful that we had pathologists within our group with the Gabor Kovacs. And even from the data, we learned a lot that many of these [00:14:00] cohorts had co pathology and many of the signs that were like Parkinson's disease, but they at the same subject maybe also signs of tauopathy. And if I remember in the clinic, we had so many questions, often, is this a real PD patient or may he develop a tauopathy? And the solution of this may be in some of these difficult patients that they have both or they develop a co pathology. And I think this is a message that we should give also our clinicians, and this makes clinical criteria especially difficult.

And the question is, should we really go for specification and even more and more specification if in some patients this is not the goal. [00:15:00] We have to consider two pathologies, so I would go for less, less single items. As Susan just said, delete some items, go for the real hard ones, like the ataxia in Parkinson's patients.

This is almost not existant. And then always think of co pathology. Maybe add one item, possible co pathology because of something like this. So keep it open because only the disease progression may show the real form and, and then in the next step, and this is something to be discussed and at a broader level and with some of our experts from the neuroscience to include biological [00:16:00] criteria and genetics.

And I think after now more than 10 years of the MDS diagnostic criteria and then, really the increasing knowledge we had in the last five years about alpha synuclein RT-QuiC tests, SAA synuclein aggregation tests. We have at least to think how we can integrate the many so technical and lab informations into some useful criteria.

It may not be available everywhere, at each site in each country, but if available it may add useful information. And I think new criteria need to be integrated, not only clinic. And even in the MDS diagnostic criteria where the olfaction is mentioned, nuclear medicine is mentioned, [00:17:00] so it's already a start, but you may have to extend that.

Dr. Eduardo de Pablo-Fernández: So, I guess it's a reflection of the evolution of the criteria. How some of these investigations become outdated or not much in use in clinical practice, and how other biomarkers or other tests can help. A proposal of this the co pathologies. I guess that's probably one of the limitations of the clinical pathological studies in general that now we are very aware of that prevalence of co pathologies in Parkinson's disease and other neurodegenerative disorders. But that was probably not the case a few decades ago. And some of these studies did not evaluate the presence of additional pathologists in detail. I think Claudia, you probably discussed this a bit, but that was my next question.

How do you fit this clinical criteria into the more broader definition of Parkinson's disease with the biological definitions recently proposed, and not just the biomarkers, but also as we discussed earlier [00:18:00] imaging markers genetics, and so on. I dunno if you have anything to add Susan, to the discussion.

What are your thoughts?

Dr. Susan Fox: Yes. I think what we found here is that the experts with their definitions of the MDS criteria back in 2015 have been mostly confirmed by what we found with pathological confirmation. So I think these clinical definitions are holding up generally, so I think we continue to use what we're doing.

But as Claudia mentioned and going forward, obviously it does need refining with modern methodologies that we hope will continue to improve our ability to diagnose a patient with Parkinson's, particularly using the multiple biomarkers that we're as a field developing within the imaging proteomics and obviously including genetics.

The co pathology portion is clearly critical. And I think the other learning from this for me really is that we still need to carry on doing clinical pathological series. We [00:19:00] still need to do that part of research, part of our clinical care if you like, going forward, it has to be continued and really hone in on these patients with multiple pathology.

And how does that refine our clinical diagnosis. But I think my take home from this is that the criteria do generally hold up for diagnosing a patient with Parkinson's disease. But obviously refining this going forward is gonna be what we'll be planning to do with modern and ongoing improvements in biomarkers.

Dr. Eduardo de Pablo-Fernández: It is a tough task to, to undertake definitely. But I think it's encouraging when you look at the clinical pathological studies, looking at diagnostic accuracy that the expert diagnosis is still probably very high top of the list and, but as you said translated that thinking process into words sometimes is a bit difficult.

But I think probably just like a simplification of these results and an update based on the new developments will be very useful in clinical practice and clinical research. [00:20:00] I dunno if you have anything to add to the discussion. What are the plans for this task in the MDS, the next steps to move this forward?

Prof. Claudia Trenkwalder: So I think the next steps would be about, this is part of the MDS, then to look at the new task force or committee finally with the mandate to establish I would say simplified version of the MDS criteria. And as we all know, many doctors just tick yes MDS criteria, but they never checked all single items because it's really too time consuming.

So I would prefer to have criteria which are really honestly checked and maybe simple. Maybe less specific, but how specific can we be? And if we add, as Susan said, [00:21:00] other techniques, biology, genetics, this may help. But we all learn during this experience how important the neuropathology is and that  we have a lot of ideas as experts and a lot of opinions, but finally, if it comes to the brain, the story may be different.

Dr. Eduardo de Pablo-Fernández: It's absolutely crucial. It's coming from me that I'm probably a bit biased, but yeah, definitely clinical pathological research is a very valuable resource and in terms of clinical diagnosis. Thank you very much Susan and Claudia for the discussion today. And thank you very much to the listeners.

I encourage everyone to read the full paper in the Parkinson's Disease Journal and bye-bye for now. [00:22:00]