Despite the arrival of FDA-approved tissue-agnostic targeted cancer therapies, there is increasing recognition that the response of tumors that are driven by the same oncogene differs according to their location.
Now, scientists at the Institute of Cancer Research in London (ICR) have shown for the first time that there are significant differences in signal transduction between KRAS-mutated cancer cell lines from a range of tumor types, and that affects how the cells respond to the same targeted drug.
The research adds nuance to the paradigm of targeted precision medicine, said Paul Workman, chief executive of ICR. "[It] shows that the story isn't quite as straightforward as matching a drug to a cancer's genetics," he said.
The study looked at 10 cell lines from each of colorectal, pancreatic and non-small-cell lung cancers (NSCLC), all of which have frequent mutations in the KRAS oncogene that acts as an off/on switch controlling cell proliferation.
The 30 cell lines were exposed to clinically relevant concentrations of seven targeted cancer drugs: AZD-5363, an AKT inhibitor; everolimus, an m-TOR inhibitor; gefitinib, which inhibits EGFR; luminespib (NVP-AUY-922), an HSP90 inhibitor; pictilisib, a PI3 kinase inhibitor; trametinib, which inhibits MEK1 and MEK2; and vemurafenib, which is targeted against BRAF.
Those seven were chosen because they are known to inhibit signaling nodes related to KRAS and/or KRAS mutations are known to affect sensitivity to the drug. Because all the drugs have been used in the clinic, it was possible to administer clinically relevant concentrations.
The scientists screened for changes in 50 phosphoproteins which are relevant to KRAS signaling and are related to targets of the drugs used as probes.
They found there were significant differences in the rewiring of signal transduction between tumor types that are driven by KRAS mutations. That influenced the response to targeted cancer drugs.
For example, a significantly lower number of colorectal cancer cells switched off AKT signaling when exposed to pictilisib, compared to NSCLC and pancreatic cancer cells. At the same time, significantly fewer NSCLC cells showed an increase in phosphorylation of MEK compared to the effect of pictilisib on colorectal and pancreatic cancer cells.
Pictilisib was discovered at ICR, and to look at response to the drug in more detail, the researchers analyzed its effect in cancer cells from 10 patients whose tumors had faults in KRAS.
The drug was shown to trigger an increase in MEK signaling in three out of seven samples from patients with colorectal cancer, but in none of three samples from lung cancer patients.
The findings, reported earlier this month in Molecular Cancer Therapeutics, suggest there are inherent differences in the signaling output in KRAS-mutated cancers. Cells from different cancers "responded quite differently" to drugs targeting the same mutations, said lead author Udai Banerji, deputy director of the drug development unit at ICR.
The research calls into question basket trials testing targeted drugs, in which any patient with a particular mutation is eligible for recruitment, regardless of the anatomical location of the cancer.
It was a basket trial involving 149 patients with microsatellite instability-high or mismatch repair deficient solid tumors that led to the FDA granting Keytruda (pembrolizumab, Merck & Co. Inc.) the first tissue-agnostic approval in May 2017. Until that point, Keytruda had been approved tumor type by tumor type. (See BioWorld, May 25, 2017.)
The FDA subsequently approved Loxo Oncology Inc.'s Vitrakvi (larotrectinib) to treat patients with any solid tumors with a TRK fusion. The drug was allowed on the market on the basis of trials in 55 patients with tumors in 17 different locations. (See BioWorld, Nov. 28, 2018.)
But Banerji said the ICR study suggests cancer type does need to be taken into account when designing clinical trials of targeted therapies. "We should no longer group people participating in clinical trials by [mutations] in their tumor alone," he said.
The research also raises the possibility that targeted drugs which failed in clinical trials of a single tumor type could be rescued in another cancer with the same oncogene. That will depend on further work to understand how targeting the same mutation can have different effects depending on the tumor type.
The research on understanding how dynamic cell signaling varies depending on tissue background feeds into ICR's broader Darwinian program, in which it aims to overcome the ability of cancers to evolve resistance to targeted drugs.
ICR scientists have developed algorithms that forecast how cancers will react when treated with a particular drug. By selecting an initial drug treatment, tumors are induced to adapt in a way that makes them highly susceptible to a second drug.
The Darwinian approach also has led to the discovery of novel Apobec (apolipoprotein B mRNA-editing enzyme) inhibitors, which are designed to reduce the mutation rate in cancer cells, slowing down evolution and delaying resistance.
The inhibitors are intended for use in combination with targeted therapies, to maintain their effectiveness for longer.