About 1,700 genes are deregulated in every single KRAS tumor, yet most are irrelevant for tumor survival. Separating mutations into three tiers shows which ones matter; in pancreatic cancer, the SMAD4 axis offers the biggest differentiation for combinations with KRAS inhibitors. Drugs that could exploit SMAD4 loss are already in clinical development for other indications.
KRAS case study, part 2 · Pancreatic carcinoma · by Pawel Zawadzki, PhD
The majority of genes are deregulated via DNA mutations, or at the level of transcription, expression, or protein modifications.
Here is shown the complexity of the molecular context in a single patient with a KRAS mutation! We all silently sweep under the rug what we cannot comprehend.
But not all gene mutations are created equal. In fact, the majority are irrelevant for tumor survival. RNAseq is noisy, but when superimposed with genetic mutations, it reveals which deregulated genes and pathways are critical for tumor survival.
If a Tier 1 mutation affects an oncogene, the job is easy. But how to exploit the vulnerability created by the inactivated tumor suppressor?
Two therapeutic strategies can be used to eliminate KRAS-driven pancreatic tumors:
Directly: p53 reactivation (Rezatapopt) is showing some early success in the clinic. Not applicable to CDKN2A and SMAD4.
Indirectly: Block MDM2 (downregulates p53), inhibitors are struggling in the clinic. Histone Deacetylase (HDAC) Inhibitors can upregulate CDKN2A, SMAD4.
Synthetic Lethal targets from CRISPR screens. Yet, no clinically validated targets from CRISPR screens.
CDKN2A is typically deleted together with MTAP. PRMT5 as a target for MTAP loss is underperforming in the clinic.
Multiple opportunities linked to SMAD4 were discovered.
Drugs that could exploit SMAD4 loss in pancreatic KRAS tumors are already in clinical development for completely different indications!
Today, this complexity is the reason behind a disappointing 6% success rate of oncology clinical trials. But this complexity can be turned against cancer.
In pancreatic cancer, exploiting SMAD4 dependencies offers unexpected opportunities for drug combinations with KRASi. Some synergistic drugs are already in late clinical development.
The complexity is scary, but it can be comprehended. Patient data + AI + smart team = better therapy
Each tissue is different. Map all tissues, then prioritize opportunities benefiting the biggest patient population.
Cell line models are typically used to design drug combinations. Here, we show that the treatment-naive patient data can identify synergistic drug combinations that will improve clinical response rates and durability.
Medicines intended to work in patients should be developed with patient data in mind.