So the APHINITY trial is a very important practice-changing study showing that if you combine two drugs that block the same molecule, HER2 versus just one drug that blocks HER2, that you have in the lymph node positive patients, so patients with cancer in the lymph nodes, you have an important benefit of the treatment. That’s what the APHINITY has shown. Now, it is also a bit of a controversial study...
So the APHINITY trial is a very important practice-changing study showing that if you combine two drugs that block the same molecule, HER2 versus just one drug that blocks HER2, that you have in the lymph node positive patients, so patients with cancer in the lymph nodes, you have an important benefit of the treatment. That’s what the APHINITY has shown. Now, it is also a bit of a controversial study. Why? Because the overall survival benefit is actually very minimal. So patients benefit good, but in terms of overall survival, do they live longer? It’s actually a minor difference. And that’s the reason why in some countries the drugs are not reimbursed. So there is an urgent need to define markers that can identify those patients who may need this double blockade versus a single blockade. And we do know from a lot of work done by the TILS working group that the immune system is a very important variable in helping to inform, and that’s an important notion to inform patients and patients about what is the outcome for me and will my tumor respond to these treatments. So what have we done? We have analyzed the immune system, the TILS, manually. I did it. 5,000 patients almost got me a divorce. We also analyzed the same cells using a computational method, but not based on artificial intelligence. We have analyzed the same cells with a computational method based on artificial intelligence, which was method developed by case 45 a startup in london what have we demonstrated that’s important for the community is in terms of prognosis and prediction both of them manual as computational derived ai versus non-derived predict outcome and benefit to the treatment. But we have also shown that if you combine, and that’s probably the way forward for AI in pathology practices, if you combine the pathologist’s work with an AI metric that identifies spatial patterns of the immune system, which pathologists see but we cannot quantify. If you combine both of them, the outcome and the benefit of the treatment is slightly better. So this suggests that in daily practice, the pathologist will continue to have a very important role in identifying the basic biomarkers like the TILs. But AI can complement and provide very important information for our patients and clinicians about having and finding those patients that benefit more from the treatment than the outcome.
