The lab’s recent findings have reinforced the message that metastasis can happen earlier than previously suspected.
One study, published in Nature Genetics, expanded on Curtis’ 2019 discovery that some tumors are “born to be bad.” She and colleagues coined the term after observing that in most patients with metastatic colorectal cancer, the cancer cells spread to distant organs, like the brain or liver, years before the initial tumor was diagnosed.
In the new paper, she and her collaborators showed that early metastasis can happen in breast and lung cancer too, two to four years before the first tumor is detected—illuminating the need for improved strategies to detect cancers earlier. What’s more, their findings illuminated the need to consider the timing and types of therapy to avoid the emergence of drug resistance.
While these are not easy messages to convey, and more research is needed, “they open our eyes to the need to study patient tumors sampled during the course of therapy, and to anticipate resistance,” Curtis says. “The seeds of metastasis can be sown early, and we need to study the continuum of disease with this in mind. It can empower the field to address some of those harder questions, particularly as new technologies are at hand.”
Recently, Curtis and colleagues also identified a long-sought-after biomarker for women with early-stage, newly diagnosed, HER2-positive breast cancer. Up to 50% of patients with HER2-positive tumors, which produce too much HER2 protein, have residual cancer following treatment.
“Many clinical trials have been designed with the goal of asking, ‘Is there a new biomarker that would predict which women would respond to therapy, above and beyond the fact that they have HER2 amplification?’ and the field has struggled,” Curtis says. “It’s unfortunate, because here’s a place where we have multiple FDA-approved drugs, and we could either combine them to increase a women’s response, or we could give targeted therapy and leave out the chemo and spare this patient the additional toxicity. So we really need predictive biomarkers.”
In a study that deployed a completely new technology called digital spatial profiling to look at 40 different tumor and immune proteins in tumor tissue samples before and during preoperative therapy, Curtis and colleagues found what they were looking for. Patients who had higher levels of an immune marker called CD45 after just a single cycle of HER2-targeted therapy tended to have a dramatic response to treatment, while those with lower levels did not. The findings were published in April 2021 in Nature Cancer.
The next step will be to design a clinical trial that tests these findings.
“If it validates, this has the potential to transform patient care,” Curtis says. “We would be able to personalize therapy with this approach—for example, by eliminating chemotherapy use in a subset of patients who do not require it.”