Opinion

Brian Shields: A very good example of the Power of AI when developed for a specific use case

Brian Shields/LinkedIn

Jan 24, 2025, 15:59

Brian Shields: A very good example of the Power of AI when developed for a specific use case

Brian Shields, Creator of KOL Pulse, shared a post by KOL Pulse on LinkedIn, adding:

“This is a very good example of the Power of AI when developed for a specific use case with curated data sets. The LAURA trial from AstraZeneca is very important for the care of EGFRmut Lung Cancer patients and has a lot of nuance based on KOL discussions.

Check out the KOL Pulse AI Chatbot results below for an example of how a model can manage very high level scientific discussions to provide insightful output.

If you’d like to try the KOL Pulse AI tool yourself, please sign up for our SABCS24 Chatbot reviewing Key Opinion Leader social discussions in Breast Cancer. Here’s the Link for the demo signup.”

Quoting KOL Pulse’s post:

“KOL Pulse AI Chat of the Day

PROMPT: provide me with a balanced debate of the clinical utility of the LAURA trial. support your debate with direct tweets from Key opinion leader sources

CHATBOT Reponse:

The LAURA trial demonstrated significant benefits for osimertinib after chemoradiotherapy in unresectable stage III EGFR-mutated NSCLC:

Median PFS: 39.1 months (osimertinib) vs 5.6 months (placebo), HR 0.16
24-month PFS rate: 65% (osimertinib) vs 13% (placebo)
CNS PFS HR: 0.17 favoring osimertinib
12-month CNS progression rate: 9% (osimertinib) vs 36% (placebo)
New brain lesions: 8% (osimertinib) vs 29% (placebo)
81% crossover from placebo to osimertinib
AE discontinuation: 13% (osimertinib) vs 5% (placebo)
Most common AE was radiation pneumonitis in both arms

Key points:

Practice-changing results with major PFS benefit
Neuroprotective effect of osimertinib
Poor outcomes in control arm, possibly due to understaging (only 55% had PET scans)
OS data immature but trending positive despite high crossover

Safety profile manageable

Multiple experts noted this redefines stage III EGFR+ NSCLC as essentially stage IV disease requiring indefinite TKI therapy. However, some raised concerns about lack of OS benefit so far and need for better staging/patient selection.

Sources:

Tom Newsom-Davis/X
Sanjay Popat/X
H. Jack West/X
Nathan A. Pennell/X
Lecia Sequist/X
Eric K. Singhi/X.”