UA-45667900-1

Tuesday, 5 August 2025

Keeping ahead of the curve in Autism (and Pitt Hopkins syndrome) treatment - the placebo effect, clinical trials, and a promising case study

Since AI is a trending tool in this blog, I decided to let ChatGPT rewrite today's post. It did rather strip out the science bits.  It added the "don't wait for permission at the end"—a little cheeky, I think. It does like to use dashes.

 

Keeping out in front of the pack is not always easy


Today’s post highlights a compelling new case study—one that turns theoretical research into a real therapy.


About time too! That was my reaction when a reader sent me the paper.

This case study reports on the repurposing of a cheap, well-known drug—Nicardipine—to treat Pitt Hopkins syndrome (PTHS). The drug had already shown promise in earlier mouse models.

So why aren’t we doing this more often? Because the system misunderstands risk.


What About the Risk?

When it comes to trying new treatments, people often fixate on the risk of the therapy itself. But that’s only half the equation. The risk of doing nothing is often much greater—especially in autism.

Most conventional drug repurposing therapies pose minimal long-term risk. Things change only when you start injecting compounds or using untested chemicals. But even then, there’s surprisingly little harm on record.

Only one death has ever been clearly attributed to a therapy for autism:

A 5-year-old autistic boy from the UK died in the US while undergoing chelation therapy. The wrong form of EDTA—disodium EDTA instead of calcium EDTA—was used. The result was fatal hypocalcemia-induced cardiac arrest. The doctor administering the therapy didn’t understand the pharmacology.

Lesson: Always read the label.

Meanwhile, the risk of death from untreated autism is well established:

  • In severe autism, common causes include drowning, accidents, and seizures.
  • In milder cases, the biggest risk is suicide.

Another overlooked danger, mentioned previously in this blog, is polydipsia—excessive water drinking—which can cause hyponatremia (low blood sodium), leading to seizures, coma, and even death.

Bottom line?


The risks from untreated autism far exceed the risks from science-based, carefully applied therapies.


The Nicardipine Case Study

A newly published study builds on promising mouse results and shows real benefit in a young child with PTHS. The drug used—Nicardipine—has been around since 1988 and is commonly prescribed to older adults for high blood pressure or angina.

🔗 Read the case study

Highlights:

  • Pitt Hopkins syndrome involves loss of function in the TCF4 gene, leading to overactivity of Nav1.8 sodium channels in neurons.
  • Nicardipine inhibits Nav1.8, making it a logical therapy.
  • In this case study, the child received oral nicardipine for 7 months (0.2–1.7 mg/kg/day).
  • Result: Mild to moderate improvement in all developmental areas, and reduced restlessness.
  • No significant side effects reported.

It’s not a magic bullet—but it’s a start.
Used as part of polytherapy, this could become a powerful tool for treating PTHS.

And there’s more coming: Vorinostat, another potential therapy, is entering human trials.


Why Don’t More Therapies Get Adopted?

A recent paper by Antonio Hardan sheds light on this. He’s the researcher who showed that the OTC antioxidant NAC benefits many with autism, and later explored the hormone vasopressin.

This time, he tackled the placebo effect—a real barrier in autism research.

🔗 Placebo Effect in Clinical Trials in Autism: Experience from a Pregnenolone Treatment Study

What They Did:

  • A two-week placebo lead-in before the main trial.
  • The drug tested was pregnenolone, a neurosteroid.
  • They used parent-reported ABC-I scores to measure irritability.

What They Found:

  • A 30% reduction in irritability—just from placebo.
  • Also improvements in lethargy, hyperactivity, and repetitive speech.
  • The placebo effect was strongest in the first two weeks, then plateaued.
  • Clinician-rated scores (CGI) did not show this placebo response.

The Takeaway:

Parent expectations strongly shape trial results—at least in the early stages.
A placebo lead-in is a clever way to measure and filter out this noise.


Early Adopters, Take Note

It pays to be ahead of the curve.

Some Pitt Hopkins parents are already trying nicardipine at home based on this case study. Good luck to them—I hope they find the right specialists and support.

Let’s not forget: the big autism trials of recent years—Bumetanide, Memantine, Balovaptan, Oxytocin, Arbaclofen—all officially “failed.”

But the drugs didn’t fail—the trial designs did.

Each of these drugs helped some individuals. The problem?
The trials weren’t structured to identify responder subgroups. We wasted time, money, and hope by not tailoring inclusion criteria more carefully.

Consider Trofinetide, the first FDA-approved drug for Rett syndrome (2023). It helps only 20% of patients, but was still approved.

I’d argue that Bumetanide has an even higher response rate in severe autism, particularly with intellectual disability—and that the best outcome measure is IQ, not a generalized autism scale.


My Own Example: No Placebo Here

How do I know I wasn’t misled by the parental placebo effect?

Simple. No one knew I was trialing treatments—not even the teachers or therapists. That meant their feedback was objective and uninfluenced by my hopes.

My son Monty went from being unable to do basic subtraction at age 9, to later passing his externally graded IGCSE high school math exam.

Not bad for a therapy that mainstream medicine still ignores.


Final Thoughts

  • Drug repurposing is safe, smart, and often effective.
  • The placebo effect is real—but it’s measurable and manageable.
  • If we want progress in autism treatment, we need smarter trial designs, not just more of them.
  • Being ahead of the curve isn’t risky—it’s essential.

💡 Stay informed, stay curious, and don’t wait for permission.


Thanks for the guest post, ChatGPT !!


One point to add to the risk assessment: by my estimation, each year in the US, around 200 to 300 people die from drowning, seizures, accidents, and suicides related to autism. In living memory, only one person has died as a result of visiting an autism doctor in the US and that death was entirely preventable.

Vorinostat, a potent HDAC inhibitor trialed in several autism models, was mentioned in the above post. Interestingly, there is a recent comment from a reader who finds it resolves 80% of his autism but only for about 2 hours. The half-life of this drug is about 2 hours. There are discussions on Reddit by people using it for autism, anxiety, PTSD etc. It is about 1,000 times more potent than HDAC inhibitors people typically might try at home. Perhaps there should be trials of micro-dosing Vorinostat? I think daily use of high-dose Vorinostat may not work well, due to side effects.  Human trials will soon inform us better. It is often older people who struggle with drug side effects, not children.  

Vorinostat may not only correct Differentially Expressed Genes (DEGs) but also:

  • Increase synaptic plasticity
  • Improve synaptic morphology (the shape and function of neuronal connections)
  • Improve memory and cognition 

The main research interest is in single gene autisms, where one specific gene is under-expressed (eg Pitt Hopkins, Rett, Fragile-X etc) but the general ideas are equally applicable to broader autism.