New model predicts chemotherapy response in colorectal cancer
Scientists may be better able to predict which patients with colorectal cancer will respond to chemotherapy using a new mathematical model that measures the amount of stress required for a cancer cell to die without harming healthy tissue.
Apoptosis, defined as programmed cell death, is believed to be a hallmark of cancer resistance to chemotherapy. The key step in apoptosis is the process that leads to mitochondrial outer membrane permeabilization (MOMP), which is controlled by different members of the BCL-2 family of proteins. Some family members promote apoptosis and some prevent it. Furthermore, those proteins that have the same effects on apoptosis work in parallel and can substitute for each other, which makes it difficult to predict whether cells are likely or unlikely to die.
For better informed decision-making in chemotherapy for colorectal cancer, Jochen Prehn, PhD, of the Royal College of Surgeons in Ireland and colleagues developed a tool that would incorporate patient-specific, molecular data sets. They studied BCL-2 proteins, determined levels of the individual proteins, and put the levels in a mathematical model that calculated what genotoxic stress level is needed to achieve apoptosis.
“Resistance of colon cancer cells in culture, as well as treatment responses of patients with stages II and III colon cancer, were critically determined by the calculated stress level required to undergo apoptosis,” Prehn said. “We found that individual patients had a high degree of heterogeneity in BCL-2 family protein levels and that this was a potential cause of the success or failure of adjuvant chemotherapy.”
Prehn and colleagues tested a clinical decision-making tool that they call DR_MOMP to determine its use in predicting treatment responses in patients with colon cancer. Using DR_MOMP, they were able to robustly predict patient outcome.
“Our study demonstrates that systems medicine approaches (ie, quantitative analysis of multiple factors in patients' samples combined with mathematical modeling) have a significant advantage over other approaches in predicting therapy responses in patients,” said Prehn. “This finding may provide a clinical decision-making tool that enables predictions of treatment responses in patients with colon cancer. As we provide a quantitative, dynamic analysis of the process of apoptosis, we can also calculate, for individual patients, the therapeutic window.”
The model could help predict how much genotoxic stress is required for a cancer cell to die before normal tissue is affected. Prehn and colleagues hope to validate DR_MOMP in other cancers and in larger patient cohorts.
“We need to develop easy and accessible protein profiling and modeling platforms that enable the implementation of this new technology in clinical trials and in pathology laboratories,” Prehn said.
This study was published in Cancer Research (2013;doi:10.1158/0008-5472.CAN-12-2269).