When doctors biopsy and treat cancer may be just as important as how they treat it. New research from Erik Herzog, PhD, the Viktor Hamburger Distinguished Professor in biology in Arts & Sciences at Washington University in St. Louis, revealed that when a treatment is given may play a key role in how well brain cancer patients respond to their chemotherapy.
Recently published in the Journal of Neuro-Oncology, the study centers around glioblastoma (GBM), an aggressive and treatment-resistant brain cancer that affects over 300,000 people worldwide every year. The standard treatment plan for this cancer is the common chemotherapy drug temozolomide (TMZ), but its effectiveness is limited. Many tumors fight back with a DNA repair enzyme called MGMT that helps the cancer cells survive TMZ's attacks.
However, the study finds that MGMT activity levels aren't constant. Both MGMT methylation - which turns the gene off - and the amount of MGMT protein fluctuate throughout the day. This led the researchers to test if the timing of tumor biopsy would also influence diagnostic results. Colleagues at WashU Medicine - Joshua Rubin, MD, PhD, Will Leidig and Omar Butt, MD, PhD - provided five years of patient biopsy data, which Herzog and his team analyzed.
"Presently, methylation is used for diagnostic purposes in cancer, particularly for identifying brain tumors with varying molecular subtypes that are correlated with differences in treatment response," Rubin explained.
This new research focused on whether methylation levels are stable or if they might change with time of day.
"We found, repeatedly, an increased likelihood of morning biopsies being scored as methylated," Herzog said. This means that taking a biopsy at different times of day could influence how doctors diagnose tumors. "We are hopeful that this chronodiagnostic approach will help identify better ways to treat this devastating disease."
These circadian cycles in MGMT gene and protein expression also regulate tumor sensitivity to TMZ over the course of a day. "When we found that TMZ is more effective in the morning, we wondered if, perhaps, this coincides with when there is less MGMT around to repair TMZ-induced damage," Herzog said.
To better understand how this timing works, Maria Gonzalez-Aponte, a graduate student in the Herzog lab, measured MGMT levels throughout the day in tumor cells and patient GBM samples. With Olivia Walch, a math biologist and co-author of the research, they created a mathematical model to predict when TMZ would be most effective relative to the daily rhythm in MGMT.
"I love when math pushes us past intuition and helps us see something new," Walch said.
The model predicted that, because TMZ takes several hours to cause DNA damage and to activate cell death, dosing TMZ right after MGMT protein peaks gives the drug the best and longest window to act while the tumor's repair mechanisms are slower.
"My main takeaway from the modeling part of this work is that, while chronomedicine sounds simple in theory, figuring out what that right time is can be tough," Walch said.
Dosing time can depend on a multitude of factors, including dose and individual variations in circadian rhythms. "The good news is that math gives us a way to chop away at that complexity and get closer to answers that intuition alone can't give us," she added.
Herzog is optimistic that this research will also one day improve the effectiveness of other drugs with similar modes of action that have known circadian targets. Beyond testing chronotherapy with TMZ, the Herzog lab is looking forward to evaluating the importance of the time of day for other signals known to promote or suppress GMB growth. This includes substances like dexamethasone, a drug used to control brain swelling. "It may be important to avoid treatment at times of day when it also promotes tumor growth," Herzog said.
Gonzalez-Aponte MF, Huang Y, Leidig, WA et al. Circadian variation in MGMT promoter methylation and expression predicts sensitivity to temozolomide in glioblastoma. J Neurooncol 176, 36 (2026). DOI
https://doi.org/10.1007/s11060-025-05242-3
This work was supported by National Institutes of Health (NIH) Grants NINDS R21NS120003 and NCI R01NS134885, and the Washington University Siteman Cancer Center. Author MFGA
