Transcription:
[00:00:00] Prof. Paul Gissen: Good afternoon or for some of you it will be morning or maybe evening. My name is Paul Gissen and I'm a professor and a consultant in pediatric metabolic diseases here in London at Great Ormond Street Hospital and University College London. And today I will be talking about the importance of recognizing the early signs of CLN2 disease. And I'm really grateful to the Excellence in Pediatric Initiative that allowed me to speak to you and I hope to get some questions from you at the end. First of all my disclosures, I have been involved in a number of clinical trials in Batten disease and but importantly all the patient images, if there are, they are all given with permission from the families. So today our learning objectives is to recall the genetics, the biochemistry and epidemiology of CLN2 type Batten disease. We will talk about the how to recognize the earlier clinical red flags, language delay, first seizures and emerging ataxias. We'll apply the clinical diagnostic algorithm, how to interpret the EEG, brain MRI scans, enzyme assays and also the gene testing. And we will discuss the effects of the current treatment with the approved drug, which is approved in many countries, cerliponase alfa and why it is important to initiate treatment early.
But I will start from talking about the CLN2 and Batten or neuronal ceroid lipofuscinosis group of disorders. It is a commonest cause of childhood dementia. There are 13 known genetic forms, there might be more. They are numbered slightly confusingly from 1 to 14, but the CLN9 is probably the same as CLN5. Most of these disorders have autosomal recessive inheritance and classically they have progressive seizures. The patients have progressive seizures, loss of motor and cognitive abilities and visual loss. The overall prevalence of the NCL disorders is approximately 1 or 2 in 100, 000. CLN2 is among the most commonly diagnosed and it is a classical late infantile, although there are some atypical forms as well. On the right, you see some of the features of the other types as well as the CLN2 and you will see that the deficiencies in heterogeneous types of proteins cause NCLs. For example, a few of them, almost half are enzymes and they are lysosomal located enzymes, but some of them are transmembrane proteins and we don't really know the exact function of many of these proteins. But today we talk about CLN2, which is a gene encoding TPP1 protein, which is a lysosomal enzyme and that stands for tripeptidyl peptidase 1. So it's an autosomal recessive condition, which means that the patients have biallelic pathogenic variants and it is a lysosomal storage disorder because it causes a deficiency of a lysosomal enzyme and there is an accumulation of the substrate for this enzyme as well as other compounds. So it's a mixture of lipids and protein, undegraded proteins and all together it's called lipofuscin and that's why the name of this group of disorders is lipofuscinosis.
[00:04:54] Prof. Paul Gissen: The onset of this disease is classically between two and four years, although the atypical forms can present later and sometimes much later. The estimated incidence of CLN2 is about 1 in 200,000 live births, so 0.5 in 100,000 live births, and there is a marked regional variation. In some ethnic groups, there is much lower incidence, and in some of them, a much more common incidence. As I already mentioned, CLN2 is phenotypically heterogeneous. So we talk mostly about the classical late infantile form, and the vast majority of patients, approximately 87%, present with this disorder, and there is quite a tight phenotype in this group of patients. The onset is between two and four years, and in the majority of these patients, it preceded by early language delay, which can be quite variable. Then, around this, between two and four, the first identifiable sign is the seizure, and that seizure could be of variable semiology, and we'll talk about that a little bit later. And then a rapid loss of motor and language function ensues. The vision loss and blindness occur slightly later, but typically still within three years or so, and untreated, these patients' life expectancy is between eight and 12 years. The atypical patients can be quite different, and also there is a variability in the phenotype of the atypical patients. Approximately 13% of total CLN2 group. The median onset is around six years, but there is quite a wide range, and some of them may even present in adulthood. Typically present with behavior or movement disorder, but sometimes seizures. They often have ataxia, dystonia, tremor, abnormal gait. The seizures may start early, but may happen later, and in some patients, they don't actually have seizures at all. The progression typically is slower, but in the majority of these patients, they still have premature death. Example of quite a later onset is the SCAR7, where there are particularly protracted phenotype.
[00:08:05] Prof. Paul Gissen: So we use a so-called CLN2 clinical rating scale that is developed both in Hamburg, but also at Weill Cornell in the US. How does it work? It's a neurodegeneration scale. It doesn't really work for patients to improve or a very fine detail of the phenotype. It has two domains, motor and language, that is being used particularly for the assessment. But there are other domains that cover seizures and visual ability. There is also other domains that are used assessing myoclonus, dystonia, and movement disorders. But simply, particularly when the clinical trial is concerned, there is this two movement and language domains that have been assessed classically. So they are scored from 0 to 3, with the 3 being the best achieved or normal function, and the 0 being a complete loss of that function, whether that's motor or language. And then we count the combined score of motor and language. The maximum would be 6, and it goes down to 0. The higher the score, the better is the function and this score is used as a standard endpoint in the natural history and in trials. Loss of more than two points on that scale is clinically meaningful decline and was used in a clinical trial as the primary endpoint of efficacy. And as I already said, it starts with motor function from normal and then when you score 3 and then it goes slightly abnormal so they can still achieve some independent gait, at least 10 steps independently, but the gait is clearly abnormal.
[00:10:25] Prof. Paul Gissen: There are falls and then when it goes down to 1 they're not able to walk independently anymore or can only crawl and then when it goes down to 0 they essentially are not able to have any independent locomotion. In language it's a little bit more complicated because, as I already said, the patients have language delay and some of them are. The delays are quite bad that they are actually non-verbal. So within non-verbal children we typically do not score the language domain, otherwise we take it the 3 as the best achieved and then with the loss of the ability it goes down to eventually complete loss of language and that scores as 0.. So the natural history of classical CLN2 disease: initially they're pre-symptomatic until the age of 2 and maybe a little bit older, but they may have some language delay. Then the early symptoms start in the region between the 2 and 4 years of age, where the language delay is obvious. They have first unprovoked seizure and soon after the first seizure there is a taxia that is progressive, the gait that is clumsy, and then with time they lose vocabulary, ambulation, the myoclonus predominates, visual failure begins and eventually there is a severe spasticity and dystonia. The patients are bed bound, essentially blind, and the premature death is inevitable without treatment. The diagnosis is a race against time because the early, early signs are non-specific.
[00:12:38] Prof. Paul Gissen: So classically there has been a 2 year diagnostic delay between the first seizure, first presentation and confirmed diagnosis and you can see that here. What we're showing you here is the graph. The y- axis is the CLN2 score and you can see that 6 is the maximum achieved and 0 is a complete loss. And the first presentation occurs around the age of 3 or maybe a little later. And then there is a rapid decline of more than 2 points in the following year or so on the CLN2 scale and the diagnosis typically happens. Unless the doctors are really astute and attuned to these problems, the diagnosis doesn't really happen until they are 5 years of age and of course the key features here is that neurodegeneration begins before the first seizure. Motor language loss is largely irreversible and what enzyme replacement therapy aims is to preserve the function as long as possible. It doesn't reverse the damage, so all that ability that is lost while the patient is undergoing various diagnostic procedures really is lost. So early diagnosis enables earlier treatment and preserving function as much as possible and, of course, testing of the siblings, looking for early detection of the siblings that are affected. So the red sign here is that there are 3 earliest signs and these are: the language delay with the unprovoked seizure manifesting, the deterioration starting and then ataxia, and the combination should definitely trigger CLN2 testing as early as possible. So, let's start with the first sign, which is the language delay. Approximately 83% of patients have variable degree of language delay. What typically happens is the first words may be spoken at the expected age of around 12 months or suddenly before 18 months, but then there is a delay in sentence formation and vocabulary acquisition. And then, once the seizure starts, there is a vocabulary regression as disease progresses. So, however, the language delay precedes seizures in this large number of children, and some of them, it is so severe that they have no words at all. And the median lag from language delay to seizure or recognition of the language delay to seizure is about 12 months. So, the second sign is the new- onset unprovoked seizures. Now, sometimes these seizures do happen as febrile seizures, so I would say be careful about these unprovoked seizures.
[00:16:28] Prof. Paul Gissen: If the patient has language delay and a febrile seizure, I would still test, although, of course, there may be more patients and the specificity of this testing perhaps less than it would be in the non- febrile seizures, but still important to diagnose. So, typically, these first seizures happen in previously well child, ages 2 to 4, and think about this language again, and this language delay being particularly mild is also possible. The semiology is heterogeneous. It could be focal. It could be generalized tonic- clonic. They could be atonic. It could be absence or could be myoclonic. It could also be drug- resistant, but not necessarily. The inadequate response to even two anti- seizure medications has been described, but it is not necessary. These patients may respond to even the first- line sodium valproate or Keppra. Myoclonic seizures are more common later on in the disease, but can happen early as well. And then the low- frequency photoparoxysmal response on EEG is a useful clue, but not always the case. So, finally, the other early sign is this ataxia and movement features. So, what do you look for there? Broad- based and unsteady gait, frequent falls, which may be a bit more than you expect for the age. They could have action tremor on reaching. They may have generalized myoclonus at rest, and the foot dystonia or toe posturing can happen, and there could be, of course, loss of recently acquired motor skills, but that means that the patient deteriorates, and you actually want to diagnose them before this happens. So, the typical progression of the movement disorder, and that slowed down by enzyme replacement therapy, is that ataxia happens before the age of four or maybe around the age of four. Then a year later, there is a myoclonus onset, and then progressive onset of spasticity, dystonia, leading to hypokinesia and chorea.
[00:19:08] Prof. Paul Gissen: So, what are these red flag combinations that you need to look for? Is that delayed sentence formation or vocabulary regression in a child who is over two? A seizure onset, new onset seizures, unprovoked, although I would be careful and febrile seizures could also be there. It may be that they are poorly responsive to first-line anti-seizure medications. What do you do? You need to send the enzyme activity assay. That can come quickly. Sequencing of CLN2 or sending a gene panel for childhood onset epilepsy is good, but it needs to be done urgently. There is no point sending this and wait for six months until that test comes back. And that's why, for me, the enzyme activity is the first-line test because typically that's easier to get. But if you have a quick turnaround for genetics, by all means, use that. Of course, a similar clinical picture could be in other patients. Some of them could be, for example, other less common Batten disease subtypes. It could be an unusual CLN1 or other variable NCLs like type 5, 6, 7, and 8. It could be a Dravet syndrome, and these patients often have febrile seizures from infancy. They could be hemiclonic, and there are genetics, of course, for that is well known now. There are other epileptic encephalopathies, and there are multiple different seizure types with a slow spike wave on EEG. It could be mitochondrial disease, typically it presents differently, but these patients may have lactic acidemia and usually multi- system features, not just neurological. It could be other lysosomal storage disorders, such as Niemann- Pick type C disease, and they have other typical features which you may be looking for, such as visceral disease manifesting with splenomegaly or typically abnormal supranuclear gaze movements. There could be ataxia and seizures related to GLUT1 deficiency. So there are, of course, other disorders that can be picked up by genetics or other investigations. So in terms of the diagnostic algorithm, you need to have clinical suspicion now unless we have screening, newborn screening, and then, of course, we don't need to think about these things, but screening is not yet available in most countries. So the first line workup for me would include the enzyme activity and genetics, which is sent urgently. But not everywhere these sort of things are available, and not everywhere there is a dried blood spot- based screening or other things. So do an EEG, and we'll talk a little bit about specific features of the EEG, but it's not diagnostic necessarily. So as I said, by chemistry, genetics, and importantly, specialist referral for these patients, whether they're going to have CLN2 or other disorders. So EEG findings in the majority of patients, there will be abnormal EEG, but it's not necessarily picked up in all of these patients. Only 41% of the patients with CLN2 will have photoparoxysmal response to low frequency photostimulation. And that's important to remember.
[00:23:39] Prof. Paul Gissen: It's not in all patients, so it's not diagnostic. And EEG and MRI are not sufficient to exclude CLN2 disease. MRI features in early disease could not be necessarily picked up by a radiologist that is not familiar with the features of this disorder. MRI findings really early, there could be nothing detectable. As the disease progresses, there will be cerebellar atrophy. Then they progress with cerebral atrophy. They may have some periventricular FLAIR on T2 and FLAIR hyperintensities, and there could be some thalamic features. But importantly, at the age of three, there may be nothing to detect. Confirming diagnosis, as I already said, it's important to do the enzyme activity. It's important to do the sequencing. And if you don't pick up two variants, pathogenic variants, and only one variant is detected, you may need to do some further analysis because there could be some deep intronic mutations. These days there are ways of targeting those. You may need to do a long read sequencing or some other more newer methods of complex detection. You may want to do the electron microscopy, but you don't need to do that if you have the enzyme diagnosis. Typically the enzyme will show less than 10% of the normal enzyme activity. And that should really be sufficient and therefore can be the first line investigation. Treatment, we talked about this, cerliponase alfa. It is a recombinant enzyme. The dose is well known. It's been approved by both FDA and EMA in 2017. And it is given in the specialist centers via Ommaya reservoir or Rickham reservoir. It's a four-hour infusion, and it needs specialist monitoring, and therefore a referral needs to happen. Briefly about the results from the pivotal trials. So the patients there, what is important to pick up, the patients in the trial were already with a progressed disease. So their age was well over the age of four. And therefore we don't expect complete preservation of function. And they also had a mean score of 3.5. So they could see they were down from six already. And important to see here, we're comparing the progression of the untreated patients in the natural history study in red compared to the patients in blue. These are the treated patients. And this is a follow up of 321 weeks and zero is the baseline. And you could see that there is a very significant slowing of the disease progression for these patients. Similarly, there was a slowing of the brain MRI, volumetric MRI loss. An MRI is a good marker of response to therapy. And what you could see here that in the first year there was a 10 percent volume loss compared to the 14 percent in the untreated patients. But after that, that loss stabilized. And this is in the whole group of patients. And I think in the earlier patients, we would have seen a less of a loss even in the first year. So there are, of course, adverse events on the trials. These are patients with recurrent seizures.
And these are various, mostly disease related rather than treatment related adverse events such as seizures, sometimes pyrexia, vomiting. There are some hypersensitivity reactions seen in the majority of patients, but they were easily overcomable with simple analgesia, sometimes using steroids to reduce the adverse events. So I think finally, the key takeaways is it's important to recognize early the language delay, seizures. It may be unprovoked. Sometimes it is provoked. It's important to test early with the enzyme activity. You may want to use it, whatever is available in your own hospital. And the gene sequencing is important, but enzyme, abnormal enzyme activity is sufficient to diagnose this patient. Treat early because that preserves the function in these patients for longer. And it's important to work as a multidisciplinary team in order for the patients to have the best outcome. So I want to say thank you to the EIP for inviting me and thank you for listening and paying attention. And I hope we'll have some questions as well.
[00:29:42] Prof. Paul Gissen: So thank you for the first question. If I am able to see it still, so how do we counsel patients particularly providing the realistic expectations? So I usually go back to the evidence and the trial results so we can see what the results show. There are now a number of publications where from ourselves also with the real- world evidence we have now patients which where we follow them up for 12 years so quite a long follow- up. What we discuss is that untreated patients have a rapid progression, loss of their skills obviously as well as loss of vision, difficult to control epilepsy. The treatment with cerliponase alfa is not a cure of the disease but what we show is the slowing of the progression of the disease, certainly preservation of the cognitive functions perhaps for longer than motor functions. But I think I go back to showing the evidence and that's probably the best. I don't promise them the complete cure but we talk about slowing of the disease progression. How often do we see children with CLN2 who present without the classic phenotype? So I said that about 13% that's the sort of overall data. I think in my own practice maybe about 20% have atypical features and they are quite different so it's important to know that they're not all the same. So some of them develop seizures quite early, some of them don't have seizures. For example I have a patient that we've been following up and from the age of two or three and he hasn't progressed at all and he is now 18 and started developing seizures. So they are quite different. So we didn't treat that patient who wasn't progressing, we followed them up with MRI scans. So I think it's important to recognize that they're not all the same. The mutations and genotype phenotype papers can be helpful. So certainly look at what the mutations are in these patients. Have you ever had a case where the enzyme activity came back borderline and then I can't see, okay, borderline rather than clearly low and how did you handle that diagnostically? That's unusual for for TPP1 to be honest with you. I think I haven't had that sort of experience. I had in many other, well in a number of other lysosomal storage disorders but not with CLN2 really. Borderline would be a difficult one. I mean we would do the same, look at the genetics, look at the electron microscopy, look at the phenotype and then yeah consider whether we would like to label this patient as having CLN2. I think it would be difficult. If it's borderline, yeah it probably isn't it. So yeah we haven't had a case like that.
[00:34:35] Prof. Paul Gissen: How often do you see a normal or near normal EEG early in the disease and does a normal EEG meaningfully reduce your suspicion? I have seen quite a bit in patients that EEG is normal and it does absolutely nothing from the point of view of to start or not to start the cerliponase alfa. So if there is a phenotype or the mutations that classical phenotype is expected or there is a sibling with a classical phenotype, I wouldn't particularly blink starting the enzyme replacement therapy if I can in a patient who has normal EEG. That wouldn't affect my decision. Are there useful biomarkers in terms of establishing prognosis or response to therapy? Prognosis, I think the mutations probably, genotype- phenotype correlation, but of course, so the majority of patients have at least one of the mutation as the common mutation. And there is quite a bit is known, but of course, when there is a combination of mutations, you can't, it's sometimes difficult to predict what the phenotype is going to be. NFL to use it as a marker of prognosis, not at this point, I think. We have done NFL and we did see it very highly raised in classical CLN2 and not as raised in patients with atypical CLN2. But to say that I'm confident to say just like that, that this is a great biomarker, I would reserve judgment at this point. I would say probably mutations is the best one. Response to therapy, that's a different point. And I think so far, there hasn't really been good evidence that there are patients who don't respond to therapy, although some particularly groups expect, perhaps their expectations a bit higher. And when they don't see this remarkable effect, they feel that the patient is not responding. I don't think that that's a marker of not responding. To me, at this point in time, probably the MRI volumetric analysis is maybe the best marker of response. But again, probably more patients need to be done to say that, and maybe a combination. So in terms of response and NFL, what we do, so we've followed our patients with NFLs. It takes a long time for NFLs to drop because NFL has a long half- life. So it could take two, three years before the NFLs come down to normal. So yes, it is a marker of response to some extent, but a very slow marker. It's not a quick marker of response. So not really at this point, there is no great markers. What is the most common misdiagnosis you see in the atypical group? Ah, God, just no diagnosis, people are baffled. So, it could be, you know, behavioral, then they see some abnormalities, so they think, what lies, because the MRI is usually done, and then they see the loss of volume, and then hopefully the investigations are done. But if MRI is not done, it could be all sorts, behavioral, it could be sitting with neurodevelopmental pediatricians and really not referred. We have a patient with atypical presentation, where her first presentation was purely with behavioral. She stopped speaking, she chose to not speak, otherwise she was fine. So, she wasn't investigated properly by neurologists for many years. And when she was, then, of course, they did the lysosomal enzymes and picked it up straight away.
So, it can't sit with neurodevelopmental pediatricians and not be diagnosed. Are there any known factors that affect how well cerliponase alfa is tolerated? So, I don't know of any predictive factors that could suggest this particular patient wouldn't tolerate. So, in my experience, there is a subgroup of patients for whom the seizures are difficult to treat, even, you know, six months a year, into a treatment with cerliponase alfa. It's a case, in most of them, it's a case of being patient and trying different medications, because, of course, we don't say to the families that the fact that we're going to give them cerliponase alfa will stop their seizures. In maybe as many as 80% of patients, what happens, they become very stable and they may not have any seizures for years. But there are some patients, and I certainly don't know the features. I mean, it's an interesting question. There are some patients that are really difficult to manage seizures. And there is also a group of patients who have motor movement disorder that is difficult to manage. Usually, these patients are later diagnosed and later starting on cerliponase alfa. So, by then, the damage occurred, and probably in basal ganglia, you know, the thalami, such that it's really difficult to treat. And it's those patients that, in my experience, stopped or were not started on cerliponase alfa because that movement disorder is so severe. And sometimes they could be misdiagnosed as something else for a long time. Has there been any evidence of an immunological response? So, there is a response. And, of course, the protein, it's a foreign protein that is administered. Compared to other enzyme replacement therapies that we use and give systemically for other disorders, it seems that the severity of the reactions, these hypersensitivity reactions, is not as high. So, there is a low severity. So far, there has not been evidence to suggest that having the antibodies sort of slows down the response to treatment. But maybe we don't have good enough biomarkers yet. But certainly, there is an immunological response. It is particularly seen peripherally, not as much in the CSF, although it can be detected also in the CSF. But to say that that has a significant effect at this point in time, we don't see that. Well, thank you very much. It was nice to see so many questions. That was really great. And so, you've been a great audience. And thanks very much. I'm going to end the webinar now.