It looks as if we’re going to get some important data on Alzheimer’s, and this time from a clinical trial that isn’t targeting amyloid. As those who have followed the field well know, tau is another protein that seems to be involved in the disease (and perhaps others as well – there’s an extensive literature on “tauopathies). Aggregated and hyperphosphorylated tau, as found in neurofibrillary tangles, is a feature of Alzheimer’s brain tissue, and these deposits are generally separate from the better-known amyloid pathology.
But the details of a tau-based mechanism for the disease are fuzzy, and it’s been a difficult thing to target clinically. Under normal conditions, tau seems to be involved in a whole list of neuronal functions and signaling pathways, and as far as I know it’s a tangle that no one has been able to sort out yet. Perhaps the most widely known attempt was the publicity some years ago about a methylene blue derivative that was supposed to decrease tau aggregation, but this ultimately led nowhere, unfortunately. I think it’s fair to say that the clinical efforts so far in this area have not shown robust “target engagement”, to use a favorite phrase in the industry. That is, they haven’t demonstrated that they’re truly hitting tau or affecting its levels, which of course makes them unlikely to have any effects if your tau-based hypothesis is worth anything. And indeed, nothing has really had any.
There is a new effort using antisense oligonucleotides (ASOs) that might break into some new territory, though. The clinical candidate (MAPTRx) binds to a region on the tau mRNA and prevents its further translation into protein. I recall the early 1990s when antisense was the hot technology; the promise of being able to tweak protein levels like this at will was tremendously exciting. Some of the technology’s boosters predicted that it could turn into the dominant mode of drug therapy, pushing aside small molecules, antibodies, and the like. But as you may have noticed, that didn’t quite happen. It proved rather difficult to make good antisense drug candidates – ones that wouldn’t get degraded too quickly, wouldn’t set off nasty immune response, could get into cells reliably, and so on.
The first antisense drug (formivirsen) made it to approval in 1998, and it treated retinitis cytomegalovirus (CMV) in HIV-infected patients. That one had the advantage of being injected into the eye, which I realize doesn’t sound like much of one. But the eye is a unique compartment, and strange molecular species can be both contained and protected from degradation there. Still, that drug only stayed on the market for a few years, because antiretroviral therapy made the CMV complication much less common. The next ASO approval wasn’t until 2013 – mipomersen for the rare genetic disease familial hypercholesterolemia. That one made the most of what happens to these sorts of chemical species when they’re injected: they go straight to the liver and pile up there, so you might as well target liver diseases like that one. Unfortunately, mipomersen treatment itself also has a significant risk of liver toxicity, so it can’t be counted as a home run, either.
There have been several approvals since then, but you’d have to say that the track record is mixed. Inotersen is approved for nerve damage in a rare hereditary amyloidosis condition, but it also has significant toxicity. There are ASO approvals for Duchenne muscular dystrophy, but their clinical benefit has (to the best of my knowledge) not been conclusively established. They were approved on rather thin biomarker data and hopes for the best, and there are a number of people (I’m one of them) who have wondered if they should have been approved at all. Short of these programs, there have been some significant ASO failures along the way as well. They’re often tried against targets like PTP1B, brought out as a last resort since nothing else had worked. And you’d have to put tau in that category, too.
This latest trial involved 46 patients in the early stages of Alzheimer’s, some of whom were double-blinded into a control group. The group getting MAPTRx had that administered intrathecally (lumbar puncture) in ascending doses, and the control patients got the same lumbar-puncture treatment but without drug. And the important result was that the higher-dose cohorts did in fact see a reduction of tau (both total and phosphorylated) in their cerebrospinal fluid over a 24-week treatment period. There really does seem to be target engagement here, and that’s a big step. A Phase II trial has started based on these results.
What that will tell us, though is still completely unknown. Target engagement is a necessary thing to prove, but it’s just the the beginning. First off, no one has ever really reduced tau before, so right up front we don’t know what will happen when you do. That goes for its possible effects on Alzheimer’s and for its other effects as well: we’re about to find out. And even if tau reduction is something that helps AD patients, we don’t know (yet) if these levels of reduction are enough to do it, or what the adverse effects might be that show up even before you get to that stage. We don’t even know yet what the effects were in this trial on tau in the brain itself – the trial has collected tau PET imaging data, but these are not yet available. You’d hope that the levels of tau in the brain tissue will track the CSF levels, but who knows?
And there are adverse effects to consider as well. There was a high level of adverse events reported in the trial, but most of these seem to have been due to repeated lumbar punctures (!) But it also appears (via MRI) that patients who got the ASO showed ventricular enlargement (which is another way of saying brain shrinkage, an effect that is of concern in the amyloid-antibody trials as well). They also showed elevated levels of neurofilament light chain protein in the CSF, which unfortunately is considered a marker for increased neurodegeneration. So there’s a nonzero chance that this treatment actually made patients worse – we just don’t know yet. As pointed out in this commentary:
Despite the difficulty in interpreting brain volume and NfL, these measures must be scrutinized as potential indicators of unintended off-target effects of the suppression of native tau and the consequences of this on axonal and synaptic function. In the context of other neurodegenerative diseases, intrathecal ASOs commonly cause an elevation in NfL, but a decrease in NfL has been observed in the few intrathecal ASO trials to report clinical benefits. For example, a decrease in NfL tracked with clinical efficacy of nusinersen, a treatment for spinal muscular atrophy that represents the first FDA-approved intrathecal ASO. Additionally, a recent FDA advisory panel concluded that a decrease in NfL is probably predictive of clinical benefit from tofersen (an ASO that targets superoxide dismutase 1) in patients with amyotrophic lateral sclerosis.
Food for thought. The rest of these Phase 1b results, and especially the eventual Phase II results, are going to mean a great deal for Alzheimer’s treatment, antisense therapies, and neurodegeneration in general. For the sake of Alzheimer’s patients and their families, I hope the news it good, but it’s anyone’s guess at this point.
