Before you invest

Why wait for Phase II data to make key decisions?

Patient data and machine learning can raise confidence in a go/no-go decision long before a readout. These lessons show how we read the record of past programs.

Investment lessons series

Eleven lessons, one method

Each lesson takes a target class, follows its clinical programs, and names the pattern that separated the ones that worked.
Start with Rule of 5.

Investment lessons Rule of 5 Five patterns behind oncology program outcomes PDFOpen → Investment lessons KRAS KRAS G12C inhibitors: KRAS is not a solo driver, so combinations will be tissue-specific. Read online + PDFOpen → Investment lessons KRAS, part 2 Pancreatic cancer: SMAD4 loss opens combination options for KRAS inhibitors. Read online + PDFOpen → Investment lessons FGFR2 A $1.9B+ gastric cancer program: FGFR2 overexpression did not make it a solo driver. Read online + PDFOpen → Investment lessons MDM2 $1B+ invested: resistance via MDM4 and hotspot TP53 mutations exists before treatment begins. Read online + PDFOpen → Investment lessons PD-L1 × VEGF Coming late is expensive: VEGF synergy with PD-L1 was visible in patient data before Phase II. Read online + PDFOpen → Investment lessons EGFR × TGFβ Second-best tissue: EGFR × TGFβ vs EGFR × LGR5 in head and neck cancer. Read online + PDFOpen → Investment lessons HER2 Why trastuzumab worked: HER2 is a solo driver in breast cancer, much less so in stomach cancer. Read online + PDFOpen → Investment lessons PIK3CA $2B+ for a PI3Kα inhibitor: PIK3CA is not a solo driver, and amplification may shape response. Read online + PDFOpen → Investment lessons EGFR Amplified EGFR is a team player: why mAbs underperformed and where bispecifics fit. Read online + PDFOpen → Investment lessons WRN The $400M algorithm mistake: a WRN mutation type missed by public algorithms. Read online + PDFOpen →

More insights are added as programs read out. Ask for a due-diligence assessment of a specific asset through the contact form.