Here is the story of a remarkable, independent treatment suggestion by GPT-5 Pro: repurposing a known drug for a patient with food protein–induced enterocolitis syndrome (FPIES).
First, how we came to test this. My close friend, physician-scientist Dr. Oral Alpan, treated the patient described in the published report (link in the thread) for refractory allergic skin disease with the biologic dupilumab, which is approved for that indication. The patient also had FPIES, a food allergy that in his case was triggered by wheat and caused hours of cramping and watery, sometimes bloody, diarrhea. There is no approved treatment for FPIES; patients are advised to avoid trigger foods and prepare for emergencies if accidental exposure occurs. For twenty years, even a small amount of wheat set off the same cascade a few hours later, followed by days of recovery.
After starting dupilumab for his skin condition, the patient traveled to France and accidentally ate a baguette. To his surprise, nothing happened! It was his first uneventful wheat exposure in two decades! On return, Dr. Alpan hypothesized that dupilumab might be responsible and supervised an oral food challenge approaching 50 grams of wheat protein. Again, no reaction! When an insurance interruption forced a pause in medication, the old symptoms returned; restarting it restored tolerance. In the peer-reviewed paper published today, Dr. Alpan and his team describe seven additional patients, ages 2 to 58, who responded to dupilumab for their FPIES condition.
While this is not definitive proof and represents an observational case series, the findings suggest that dupilumab could be a potential treatment for FPIES. Dr. Alpan intends to contact Regeneron, the manufacturer, to pursue clinical studies hopefully toward FDA approval for this disease.
About two months ago, as he had just submitted the paper on this case, Dr. Alpan told me this story. Until today’s publication there were no reports in the literature of dupilumab potentially treating FPIES. Around that time GPT-5 Pro had just been released, and I thought this would be a perfect case to test whether AI could infer a treatment from the clinical case alone. I remember anxiously watching it think for about 12 minutes. When the output appeared, I almost fell out of my chair! Its top recommendation read: “ dupilumab (IL-4Rα blocker).” (see attached screenshot). Just like my friend had done, it was leveraging the skin condition to treat FPIES. A key insight could be that GPT-5 Pro had recognized this drug is not only effective for the skin condition but also has a “broader epithelial effect,” while this has not been proven in FPIES patients, I think it is a plausible mechanism in this disease. It also listed other options that were mostly symptomatic and flagged one with an unfavorable side-effect profile. There were several other suggestions, though those are most obvious ones and also suggested by other models too. No other model I tried identified dupilumab as a first choice; though GPT-5 Thinking suggested it as a second or third options.
Many other treatments for currently untreatable conditions may already exist and can be uncovered by using GPT-5 Pro and other advanced AI models to repurpose drugs (among 20,000 FDA approved drugs!). Of course these will still need to be clinically tested but this is how AI can help catalyze a healthcare revolution! I am very excited about what comes next and future AI models will be even more powerful to make such discoveries!
Important safety and regulatory information:
Dupilumab is not approved by the U.S. FDA for the treatment of FPIES. In this study, dupilumab was used on-label to treat each participant’s comorbid, FDA-approved allergic indication; the favorable FPIES observations occurred in that clinical context and should be considered off-label with respect to FPIES. Patients should not initiate or change therapy without consulting their physicians. This press release is for informational purposes only and is not medical advice.
Also, as evidence that this was done before today’s publication, I sent a GPT-5 Pro link to friends at
@OpenAI about two weeks ago to give them a heads-up, but I waited until the paper went online today to reveal this story.