This was an excellent, clinician-grade walkthrough of what “success” should actually mean in FTD-GRN: not just moving a biomarker, but stabilizing the neurodegeneration trajectory (NfL/atrophy) early enough to preserve function, because we’re not regrowing neurons, we’re trying to stop the slide.
I also appreciate how clearly you surface the real tradeoffs in the current approaches: AAV gene therapy is conceptually elegant (fix the root cause), but the body’s immune system + systemic biodistribution are not trivial, and “one-time” only counts as a win if the safety profile is truly clean. Meanwhile the indirect strategies (e.g., manipulating trafficking/uptake) can raise plasma markers yet still miss the core intracellular lysosomal biology.
Your point about timing is the one I keep coming back to: the therapeutic window is likely prodromal/very early, before secondary toxic cascades become self-sustaining.
Curious how you think about the clinical trigger to treat in practice: genotype + subtle change, imaging, rising NfL, or a composite?
Thanks for the comments! I also have been thinking a lot of the therapeutic window. That is one of the main lessons I take from the results of the trials in Alzheimer's disease. The earlier in the time course of the disease we can intervene, the better a treatment will perform.
Passage is recruiting NACC FTLD GS 0 patients now which is the most reasonable attempt at intervening as early as possible. Of course, the "best" time would be to intervene in early childhood so that we restore brain chemistry to normalcy (or so we think) as early as possible
In practice, from a clinical perspective, I favor sequencing as early as possible to understand if FTD-GRN is in the cards in the future for the patient. Then, I think shared decision making with the patient as to WHEN they would like to initiate treatment where the clinician explains based on natural history data the time course and onset of symptoms. So treatment based on genotype + (patient decides when OR NACC FTLD GS hits 1-2, whichever comes first).
NfL lacks specificity, imaging is costly and not revealing in early disease as far as I understand.
interesting to see LLY add another arm: mild disease (as of december). Consistent with your early intervention theory for disease modification. It's global score .5,1 still. Just lower on the CDR Sum Boxes (from < 15 to < 10).
No. Sequencing + shared decision making with patient OR development of early symptoms I think is most likely. CSF NfL is invasive and nonspecific and in my opinion, offers little value diagnostically.
However, once given treatment or for monitoring purposes, NfL may be useful IF the patient wants repeat CSF sampling.
good point. PGRN systemically is associated with a cancer signal if I remember correctly. But, how long does it take for this to manifest? And what is the effect size? Certainly something worth thinking about. I suppose plasma PGRN is the relevant marker here for systemic exposure.
We may need to wait for the development of superior vectors (relative to AAV9) as a prerequisite to generating unequivocal clinical data. There are at least a couple of public companies who may be relevant in this regard (Sangamo being one).
Alector (INFRONT-3) failed pretty completely so the indirect Sortilin approach is clearly not valid.
Also, given Catalent’s challenges, I’d be a bit concerned relying on them as a sole CDMO.
This was an excellent, clinician-grade walkthrough of what “success” should actually mean in FTD-GRN: not just moving a biomarker, but stabilizing the neurodegeneration trajectory (NfL/atrophy) early enough to preserve function, because we’re not regrowing neurons, we’re trying to stop the slide.
I also appreciate how clearly you surface the real tradeoffs in the current approaches: AAV gene therapy is conceptually elegant (fix the root cause), but the body’s immune system + systemic biodistribution are not trivial, and “one-time” only counts as a win if the safety profile is truly clean. Meanwhile the indirect strategies (e.g., manipulating trafficking/uptake) can raise plasma markers yet still miss the core intracellular lysosomal biology.
Your point about timing is the one I keep coming back to: the therapeutic window is likely prodromal/very early, before secondary toxic cascades become self-sustaining.
Curious how you think about the clinical trigger to treat in practice: genotype + subtle change, imaging, rising NfL, or a composite?
Thank you!
Thanks for the comments! I also have been thinking a lot of the therapeutic window. That is one of the main lessons I take from the results of the trials in Alzheimer's disease. The earlier in the time course of the disease we can intervene, the better a treatment will perform.
Passage is recruiting NACC FTLD GS 0 patients now which is the most reasonable attempt at intervening as early as possible. Of course, the "best" time would be to intervene in early childhood so that we restore brain chemistry to normalcy (or so we think) as early as possible
In practice, from a clinical perspective, I favor sequencing as early as possible to understand if FTD-GRN is in the cards in the future for the patient. Then, I think shared decision making with the patient as to WHEN they would like to initiate treatment where the clinician explains based on natural history data the time course and onset of symptoms. So treatment based on genotype + (patient decides when OR NACC FTLD GS hits 1-2, whichever comes first).
NfL lacks specificity, imaging is costly and not revealing in early disease as far as I understand.
https://clinicaltrials.gov/study/NCT04408625?tab=history&a=16&b=17#version-content-panel
interesting to see LLY add another arm: mild disease (as of december). Consistent with your early intervention theory for disease modification. It's global score .5,1 still. Just lower on the CDR Sum Boxes (from < 15 to < 10).
I wonder if LLY says anything in Q4 Call.
Great piece! Do you think biomarkers like NfL will end up being the key trigger for when to treat?
No. Sequencing + shared decision making with patient OR development of early symptoms I think is most likely. CSF NfL is invasive and nonspecific and in my opinion, offers little value diagnostically.
However, once given treatment or for monitoring purposes, NfL may be useful IF the patient wants repeat CSF sampling.
Do we thing all that excess PGRN is an issue?
good point. PGRN systemically is associated with a cancer signal if I remember correctly. But, how long does it take for this to manifest? And what is the effect size? Certainly something worth thinking about. I suppose plasma PGRN is the relevant marker here for systemic exposure.
We may need to wait for the development of superior vectors (relative to AAV9) as a prerequisite to generating unequivocal clinical data. There are at least a couple of public companies who may be relevant in this regard (Sangamo being one).
Alector (INFRONT-3) failed pretty completely so the indirect Sortilin approach is clearly not valid.
Also, given Catalent’s challenges, I’d be a bit concerned relying on them as a sole CDMO.