Monthly Archives: March 2024

How much PSP is “important”?

Posted on by

I could use your input right now.  (Actually, I could always use your input, but only occasionally do I ever specifically ask for it.)

A few days ago, I attended a two-day conference in Washington, DC on the tau protein sponsored jointly by the Alzheimer’s Association, The Rainwater Foundation and CurePSP.  A talk on clinical trials in the non-Alzheimer tau disorders mentioned the well-known difficulties in recruiting adequate numbers of patients with rare conditions like PSP.  In the Q/A, I asked if there’s a realistic possibility of reducing the number of patients required for a trial by using a new approach called “personalized endpoints.”  Afterwards, the editor of a journal introduced himself and asked me to write a review article/opinion piece on that issue. I said OK, but now I could use your help.

Here’s the background to my question at the conference, though many of you already know what’s in the next two paragraphs:

The typical Phase II or III clinical trial divides the patients into active treatment and placebo groups.  Trials of chronic, progressive disorders like PSP measure the signs and symptoms for each patient at “baseline,” i.e., the first visit after the screening visit, using a battery of scales and tests.  One of those, which for PSP, almost always the PSP Rating Scale, is deemed the “primary outcome measure.” Other measures of the drug’s effect are called “secondary” outcomes and still others under evaluation for future use are called “exploratory” outcomes. 

At the end of the double-blind period, typically one year for PSP, the battery of outcome measures is repeated, many of them having been repeated at interim visits as well.  Then, for each treatment group, disease progression is measured as the rate of progression (i.e., PSPRS points per year) from baseline to endpoint.  If that difference is less, on average, for the active group than for the placebo group, and is large enough that the likelihood of having occurred by chance is less than 5 percent, then the result is deemed “statistically significant.” If that result is reinforced by similar results in at least some of the secondary outcome measures and if the side effects are justified by the efficacy given the disorder’s severity and availability of other treatments, the drug will then be considered for approval by government regulators.

So what’s the problem?   

First of all, “statistically significant” is not the same as “clinically significant.”  That means that a result too small to make a difference to the individual patient can, because of a study’s large size, reach statistical significance.  The FDA knows this, of course, and relies on secondary outcome measures to verify clinical significance.  But the secondary measures may under-perform statistically, or may measure only a single aspect of the disease such as cognition or balance, or may lie far from the patient’s lived experience, as do, for example, an MRI or a blood test.

Secondly, averaging the entire active drug group and the entire placebo group is a very coarse measure.  That means that demonstrating a given treatment effect with statistical significance requires large numbers of patients and/or a longer study.  Both issues mean more expense for drug companies, which means that fewer drugs will get tested, the trials will be longer, and effective drugs may appear to be ineffective (called a false-negative result or Type 2 error).  None of those things is in anyone’s best interest.

So what’s the solution?

The PSP Rating Scale is far from perfect and various improvements have been published.  While each improves upon the original in some way, none is more sensitive to change.  The only outcome measure confirmed to be more sensitive to change than the PSPRS is the MRI-based measurement of brain atrophy, and that’s too far removed from actual symptoms and disability.  So, we need new study designs that can squeeze more information out of fewer patients.

A more sensitive way to assess a drug’s benefit uses “personalized outcomes.”  That’s where each patient being enrolled is assigned an expected endpoint PSP Rating Scale score based on how much they are likely to progress over the following year according to published research.  Relevant baseline data includes things like age, sex, progression since onset, baseline PSPRS score and subsets thereof, certain MRI abnormalities, and levels of certain chemical markers in the spinal fluid or blood.  At the end of the double-blind period, the active drug and placebo groups are compared with regard to how many patients did better than their own pre-defined expectation.

But how much better is “better”?  Enter a concept called “minimal clinically important difference.” That’s exactly what it says: The smallest change in a test score that corresponds to a change that makes a difference to the patient or family.  The trick is how to obtain this information in some sort of reliable, standardized way.  For PSP, the only attempt to do this to my knowledge was published in 2016 by Dr. Sarah Hewer and colleagues, mostly from Alfred Hospital in Melbourne, Australia.  They mined data from a completed, negative PSP trial that used a secondary outcome measure widely employed in clinical research, the “Clinical Global Impression of Change” scale (CGI-c).  This simple, seven-point scale asks the study neurologist to decide if overall, compared to baseline, the patient is “very much,” “much,” or “minimally” improved; or unchanged; or “minimally, much, or very much” worse.  Hewer et al, calculated the average degree of PSPRS worsening for patients rated by the CGI-c as “minimally worse” over the course of the year of the trial.  (Of course, the CGI-c uses the neurologist’s opinion, not the patient’s or family’s, but hopefully, the neurologist relied on their input.  I certainly always did whenever I completed a CGI-c.)

The minimal clinically important difference in the 100-point PSPRS turned out to be 5.7 points, with a 95% confidence interval (the range encompassing 95% of the patients) of 4.83–6.51.  Now, 5.7 points on the PSPRS represents about six months’ decline for the average patient with PSP-Richardson’s syndrome, with most of the other subtypes declining a little more slowly. 

Finally, here’s my question for you: 

As the seven steps in the CGI-c may be too coarse or too fine, is a six-month decline really the “minimal” worsening that’s important to you?  Or would a smaller or larger decline be the least you’d consider important?  You or a helper can judge this in terms of your overall comfort, or your ability to perform daily activities, or a combination of the two.  Note that I’m asking how many months’-worth of decline is important, not merely noticeable, which I assume would be less.

Please respond in the comments feature or by email: ligolbe@gmail.com.  Thank you!

An encouraging drug trial, but caveat emptor

Posted on by

A bit over two years ago, my posts from December 19, 2021 and January 22, 2022 reported that the drug censavudine (TPN-101) was entering a Phase 2a clinical trial. Now it’s over and the results were encouraging.  I choose that word carefully, and here’s why:

The drug company sponsor, Transposon Therapeutics, presented the results at a conference in Lisbon a couple of weeks ago and I’ve only now had the opportunity to see the numbers. The trial was small (42 patients, of whom 10 were on placebo) and brief (a bit less than 6 months for the double-blind phase), with an open-label extension of the same duration.  This study design is typical of Phase 2a trials in neurodegenerative diseases, which are designed to study safety and tolerability because their small size and short duration can’t detect clinically meaningful benefits.

The company’s press release says, “Participants treated with TPN-101 for the entire 48-week trial duration showed a stabilization of their clinical symptoms as measured by the PSP Rating Scale (PSPRS) between weeks 24 and 48. In contrast, participants who had been on placebo from weeks 1 to 24 continued to show a worsening of the PSPRS between weeks 24 and 48, suggesting a delay of clinical benefit onset of at least 24 weeks after start of drug treatment, and lagging behind the early effects on biomarkers seen in weeks 1 to 24.”

Here’s an image from the poster presented at the conference. (Sorry about the blurriness – it’s a screen shot.)

The vertical axis shows the number of points, positive or negative, by which the PSP Rating Scale changed from baseline.  (Up is worse. The typical progression of PSP is 10 or 11 points per year.) The purple lines, both dashed and solid, are the patients assigned to placebo for the first 24 weeks and the other colors are the groups on the three different dosages of censavudine (100 mg, 200 mg and 400 mg per day).  The brown dashed line is the aggregate of the three.  The vertical line segments with hashmarks show the scores’ standard errors, a measure of variation among the patients. (See my P.S. at the end of this post.)

Note that by the end of the 24-week double-blind phase, all the results are more-or-less superimposed, meaning that the patients on active drug did no better than those on placebo.  That’s normally considered a negative outcome in terms of neuroprotection, but of course this trial was too small to reveal a neuroprotective effect of realistic magnitude.  After the placebo patients started receiving active censavudine at Week 24, they continued approximately along the same progressive path with the usual random wiggles (dashed and thick purple lines), ending at about 10 points worse than baseline at Week 48.  This also would ordinarily be interpreted as an absence of benefit. 

Now, here’s where things get tricky:  After Week 32, which was eight weeks into the open-label phase, the patients on censavudine since baseline seemed to stabilize – that’s the flattening of the thick green line.  In other words, their rate of progression from Week 32 to Week 48 was much less steep – a point or two rather than the expected three or four points.  So, looking at the small magnitude of progression just from Week 32 to Week 48, Transposon suggested that censavudine may have a delayed benefit that became evident only after 32 weeks of treatment.  In this view, the placebo group, having started censavudine at Week 24, hadn’t had enough time by Week 48 for the benefit to express itself. 

Here’s the problem with that reasoning:  Looking at just two visits (Weeks 32 and 48) doesn’t provide enough statistical power to conclude anything, especially with so few patients (30 patients on censavudine from the start) and only 16 weeks to work with.  Furthermore, it’s easy to see that the censavudine and placebo groups, with some random wiggling, both follow the same line not only from baseline to Week 24, but also from baseline to Week 48 and from Week 24 to Week 48. 

The PSPRS results were not statistically significant, of course, and Transposon didn’t claim they were.  However, their presentation’s conclusions did say, “Clinical improvements emerge with longer treatment.”  I say, “There’s insufficient evidence for that.”

The trial included spinal fluid sampling at baseline, Week 24 and Week 48.  They tested for several things, including interleukin (IL)-6, which correlates with brain inflammation, and neurofilament light chain (NfL), which correlates with loss of brain tissue in PSP.  Here are the IL-6 results:

From baseline (“W0”) to Week 24, IL-6 for the group on the highest dosage of censavudine declined and continued to do so for the next 24 weeks of the same dosage.  But for the groups on placebo and lower dosages of censavudine during the first 24 weeks, the IL-6 levels rose and then, after switching to censavudine (at the highest dosage), their IL-6 declined to become indistinguishable from that of the group on high-dosage censavudine from the start.

This looks very good, though I’m not sure why the placebo patients were able to “make up for lost time” so nicely.  In any case, the result was not statistically significant, but at least it’s encouraging.  

Now, let’s consider the NfL results, as shown in the graph below.  In the group on 400 mg throughout, the level showed no change over the whole 48 weeks.  By Week 24, the patients on placebo worsened to an extent predicted by previous research and then seemed to improve, or at least to stabilize, after starting to receive censavudine 400 mg.  This result, like that of the IL-6, was not statistically significant, however.  The 100 mg and 200 mg groups, upon switching to 400 mg at Week 24, seemed to accelerate in their degenerative course, a puzzling result.  In any case, none of these observations reach statistical significance.

Despite the lack of statistical significance for the IL-6 and NfL results, Transposon’s bottom-line conclusions read, “Reductions in NfL and IL-6 support the potential for a positive effect on neuroinflammation and neurodegeneration.”  Technically speaking, that nuanced statement is appropriate, as they do “support the potential,” but the very large titles for the two graphs above were the very non-nuanced “TPN-101 Reduces Neuroinflammation” and “TPN-101 Reduces Neurodegeneration.”  I didn’t include them in my two screen shots because I didn’t want to perpetuate those very misleading declarations.

My bottom line: Censavudine (TPN-101) is acceptably safe over two years’ observation.  This trial was too small and its double-blind phase too brief for any statement as to efficacy in slowing PSP’s progression.  A large trial able to demonstrate whatever efficacy may exist is well justified and I look forward to it.

My other bottom line: Transposon’s presentation of its study’s efficacy results is misleading at best.

—

P.S. for the statistically interested:  This presentation used standard errors (SE) rather than standard deviations (SD).  SD is appropriate when two groups such as a placebo group and an active drug group are being compared.  But for tracking the course over time of a single group, SE is appropriate.  This presentation includes both kinds of observations, so I can’t fault their choice of SE.  But you should keep in mind that the SE is a smaller number than the SD:  SE is the SD divided by the square root of the N.  So, in this trial, the N of each dosage group and the placebo group was about 10, and the square root of 10 is about 3, so the height of the error bars in each graph is only about a third of what most of us are used to looking at.  That creates a false impression of a meaningful separation between the placebo and censavudine results.

A correction

Posted on by

A few days ago, I posted something about a clinical trial of a monoclonal antibody still accepting enrollees in Europe. I’ve since discovered that that’s incorrect, and that trial is filled. Sorry for the error. I’ve taken down that post.

But several other trials are coming down the pipeline in Europe, North America and elsewhere in the next few months. I’ll keep you informed (best I can).

Good news from the FDA

Posted on by

Another step forward: The FDA, for the first time, granted permission for a tau PET ligand to proceed to a Phase 3 clinical trial.  What does that mean?

For background, please see my post from November 3, 2021, called “OK, so maybe we do have a marker.” 

A compound called “fluorine-18 APN-1607” (18F-APN-1607) is being developed as a PET ligand for PSP because of its affinity for the tau protein.  It specifically targets 4-repeat tau, the type found in the neurofibrillary tangles of PSP and corticobasal degeneration.  The ligand, which we’ll call simply “1607,” has passed safety testing and has been found able to distinguish PSP from its most likely diagnostic alternatives: Parkinson’s disease and the Parkinsonian variant of multiple system atrophy.  Of course, both of those are alpha-synucleinopathies, not tauopathies.  But 1607 can also distinguish PSP-RS from the two next-most-common kinds of PSP, called PSP-Parkinsonism and PSP-progressive gait freezing, and from corticobasal degeneration. 

Now the challenge is to see if 1607 can do those things better than a good neurologist, which means also a useful degree of accuracy in the early, diagnostically equivocal stages, not just after the disease is well-established, as in the trials to date .  The FDA-approved plan for the new trial will administer the scan to people with “possible” PSP, wait up to 18 months, and then compare those with positive scans to those with negative scans with regard to whether they’ve progressed to “probable” PSP.  The formal PSP criteria published in 2017 provide formal definitions of “possible” and “probable” PSP.  “Definite” PSP requires autopsy, and as part of its statistical analysis, the trial will analyze any patients, hopefully a minority, who come to autopsy during the 18-month period.

How long will this all require?  It will probably take a year to recruit all the patients, and the last patient will finish 18 months later.  Add a few months at the start for administrative issues and a few more at the end for data analysis, and we’re talking about 3 years.  But it’s progress, and the FDA has indicated that they would consider approving the product on the basis of this single Phase 3 trial, assuming the results are convincing. For most drugs, the FDA requires more than one Phase 3 trial.

Disclosure: The drug company behind 1607 is called Aprinoia Therapeutics, based in Cambridge, Massachusetts.  I serve as a paid consultant for them, advising in design of their trial.  I have no stock in Aprinoia or other financial interest in the success or failure of 1607.  In that spirit, I’ll also mention that another tau PET ligand, 18PI-2620, sponsored by a German company called Life Molecular Imaging, is close behind 1607 in the development pipeline and is showing similar utility.

No panacea

Posted on by

In case you missed the front-page story three days ago (3/8/24) in The New York Times, the drug Relyvrio has failed to show benefit in a large (664 participants), Phase 3 trial in amyotrophic lateral sclerosis.  The drug did appear to show benefit in a much smaller (137 participants) Phase 2 ALS trial in 2020 and was provisionally approved for sale by the FDA on the strength of that result.  Now, the drug company, Amylyx, may have to discontinue marketing the drug for ALS. 

Why is this relevant to PSP?  Because four months ago Amylyx initiated a 600-participant Phase 3 trial of Relyvrio for PSP.  It’s called “ORION.” So far, recruitment has begun at only a handful of sites, all in the US, with plans to expand into Europe and Japan over the coming months.  The FDA’s permission to start ORION without a Phase 1 or 2 in PSP was based in part on the success of the drug in the Phase 2 ALS trial.  See my post of February 29 for details. 

Amylyx is also testing Relyvrio in people with Alzheimer’s disease, where a Phase 2 trial has demonstrated adequate safety and tolerability. I have no information on a Phase 3 in AD.

The question now is whether the ORION trial in PSP will continue.  So far, there’s been only one business day since the ALS news, and I’m not sure if the top brass at Amylyx — or the company’s sources of financing — have yet decided.  But the Times article reported that the FDA approved Relyvrio for ALS only after Amylyx agreed to withdraw the drug if the Phase 3 trial showed no benefit.  Furthermore, right after the ALS trial news hit, the stock price of Amylyx dropped from $19 to $3 and stayed about there.  Stock markets usually know how this sort of news is likely to play out.  Relyvrio is Amylyx’s only marketed product but they do have other drugs in the development pipeline.

The mechanism of action of Relyvrio addresses issues important to both disorders, which suggests that if it failed in one, it could well fail in the other.  But we don’t really understand the pathogenesis of either disease well enough to know if PSP might respond when ALS did not.

Meanwhile, if you were planning to try to enroll in the ORION trial, I’d advise you not to change your plans.  The ALS trial showed no important toxicity, at least in people with ALS, and you wouldn’t want to lose your potential spot at the study site because you delayed enrolling until definite information on the future of the ORION trial became available.  When other trials in PSP start enrolling, that advice could change, of course.

(I’ll repeat the disclosure I made in my 3/8/24 post: I’m a paid consultant for Amylyx, advising them on design of the ORION trial and training the participating neurologists on proper administration of the PSP Rating Scale.  But I have no stock in the company nor other financial interest in the drug’s success.)

Bad company

Posted on by

I have good news and bad news, and they’re the same piece of news.

You’ve heard of Huntington’s disease.  That’s the strongly hereditary condition that causes dementia and involuntary movements, starting at an average age of about 40 and leading to severe disability and dependency within 15 years or so.  Its most famous victim is Woody Guthrie, the folksinger/songwriter who died in 1967.  HD caused by a pathogenic variant (the word “mutation” is no longer considered appropriate) in the gene encoding the protein “huntingtin,” rendering it toxic to brain cells.

The variant in HD is a “trinucleotide repeat expansion.”   The gene has a region where the three nucleotides cytosine, adenine and guanine (CAG) are repeated multiple times.  The normal number of such repeats is less than 27.  Everyone with 40 or more repeats will develop HD and each of their children has a 50% chance of inheriting the abnormal version of the gene.  For 28 to 39 repeats, the person usually remains healthy, but the number of expansions can increase during the process of producing sperm cells or ova, putting subsequent generations at risk.  A number of other neurological diseases are caused by CAG expansions in different genes.

Huntingtin protein has several known functions in the brain, but the damage to brain cells probably is not the result of loss of any of those.  Rather, the excess number of CAG repeats encodes a an excessively long string of the amino acid glutamines. That gives the protein its toxic property, but we’re still not sure how it works or what to do about it. We do know, though, that in HD, brain cells have clumps of strands of glutamine, just as in PSP there are clumps of tau protein.

In PSP, there is no single culprit gene.  Unlike HD, PSP only very rarely runs in families and each of the 14 genetic alterations currently known to be associated with PSP confers only a slight degree of risk.  I know this looks cryptic, but here’s the current list of PSP-related genes in approximate order of discovery date: MAPT, STX6, MOBP, EIF2AK3, SLCO1A2 (two loci), DUSP10, RUNX2, LRRK2, APOE2/2 genotype, CXCR4, EGFR, GLDC, and C4A.

To this list we can now add “HTT,” the Huntington’s gene.  That’s because last month, researchers announced that they had sequenced HTT from 588 people with autopsy-proven tauopathies, including 98 with PSP, along with controls without neurodegenerative disease.  They found that while only 0.2% of the controls had at least 27 CAG repeats, for the people with PSP, the figure was 3.2% and for those with corticobasal degeneration, 2.7%.  Keep in mind that in only a small fraction of those 3.2% will the number of CAG repeats expand far enough into the toxic range to pose a risk to the children or grandchildren. 

The paper’s first author is Dr. Sergio Pérez-Oliveira and the senior author is Dr. Victoria Álvarez, both geneticists at Hospital Universitario Central de Asturias, in Oviedo, Spain, along with 21 other collaborators.  The paper appeared in Brain Pathology, a well-respected journal.

So, the bad news I mentioned is a very slight increased risk of HD occurring in future generations of families with a member with PSP.  What’s the good news?  It’s that we now know of another genetic risk factor for PSP, and it’s one that we already knew a lot about, thanks to decades of research on HD.  We can compare the long list of known actions of huntingtin in the brain to the long list of actions of the 14 other PSP-related genes. More important, we can compare the known toxic action of the excessively long strings of glutamine with the list of ways in which brain cells are damaged in PSP and look there for synergistic interactions and for drug targets to disrupt such processes.

Overall, I’d say that the good news definitely outweighs the bad news, first because statistically, the bad news is only a very small risk, and the new publication doesn’t change it – it only reveals it.  But the revelation of the good news gives PSP researchers something new to sink their teeth into in their search for a prevention and cure.

I’m glad you asked that . . .

Posted on by

Yesterday a reader left a comment regarding my 2/29/24 post on the ORION trial (of the drug AMX-0035) and I responded on the comments page. But I thought the comment was so well expressed and possibly so widely shared by my readers that it deserved more of a platform. So I turned the question and my response into this post.

Hello Dr. Golbe,

Thank you for the information and data on the AMX-0035 trial. I cannot help but share my thoughts.

The endpoint of this trial, if I understand correctly, is to slow the progression of the disease’s natural course. How is this to be assessed via the PSPRS? An absence of a rise in the score over time or an improvement in the score? If the goal is to prevent a rise in the score over time, is there a known rate at which the PSP-RS usually rises in the absence of any intervention in order to compare this to?

I hesitate to bring this next concern up, however I feel that I need to. In the absence of any known treatment for PSP, if this trial is successful then this is quite good news – anything that helps in any way is good news. However, is it really? The quality of life for a patient with PSP is terrible as you (or anyone who has ever cared for or evaluated a patient or loved one with PSP) know. Therefore is extending this poor quality of life by a year truly a success? The reason I bring this up is because I would like to know if perhaps by targeting these molecular and cellular processes within the mitochondria and endoplasmic reticulum, and thus reducing the stress and burden which is on these patients at the cellular level, is there any hope that perhaps there will also be some symptom improvement as a result of a lessening of burden/overload of the system and the brain’s own immune system and other processes being able to more efficiently function or discard of more abnormal proteins? And therefore some (even small) improvement in quality of life. This very well may be completely unknown. I may have asked you this in a prior post. I ask this not to be dismal or morbid but to see if there is even more hope. As you always say, hope is important.

AF

Dear Ms. F,

I think both of your questions are shared by many others.

The answer to you first question, regarding what the patients on the study drug are compared to, is the placebo group. At the time of enrollment, each patient is randomly assigned to receive either the real study drug or an identical-appearing placebo. Of course, this plan is made quite clear in the informed consent process but only the drug company knows the assignments. For the ORION study, 60% will get real drug and 40% placebo.  At the end of the study, the rates of progression of the PSPRS score for the two groups are compared. In this way, we don’t need to know in advance how rapidly PSP progresses. On completing their 12 months of placebo or real drug, each patient will be offered the chance to take the real drug (called the “open-label phase”).

The second question asks about the likelihood of symptomatic benefit, rather than just slowing the rate of worsening. The chance of that is low, but not zero. If the drugs do improve the function of cells affected by the PSP process, some of them may be able to recover but others (probably most) will be beyond saving. Either way, the treatment may allow cells that are still healthy to avoid becoming involved in the disease process at all, or with a major delay. What we don’t yet have is a way to restore the function of the cells already lost, though researchers are busily working on that.

Your second question implies that it may not be worthwhile merely to slow the progression of a disabling condition without curing it or improving its symptoms relative to the study baseline. That’s a legitimate philosophical and ethical question. My own interactions with patients in my decades of practice and my informal poll of this blog’s readers show that even a 25% slowing of progression (the benefit of AMX-0035 in ALS), which would provide one more year of life for people with PSP, would be worth the hassle and risk of side effects of a new drug. Keep in mind that in prolonging survival from three years to four, the drug would not merely prolong the most advanced, disabled stage for another year. Rather, it would prolong the condition in each of the three years by 33%.

I hope this clarifies things a bit.

Dr. Golbe