About 1 in 8 women in the United States develop breast cancer. Although early detection efforts have reduced mortality, treatment options remain complex. While this is partly due to the heterogeneity of the disease, in recent years we have seen a significant increase in the number of therapeutics approved to treat cancer patients. This adds to the complexity of determining which patients will benefit most from selected treatments.
Given these challenges, sophisticated and accurate predictive software tools are required to incorporate precision cancer care to better guide and support clinical decision-making. Technology and artificial intelligence, in particular, have the potential to improve clinical decision-making, thereby transforming the lives of patients.
Histologically, breast cancers are classified according to the presence or absence of molecular markers, such as estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2). These molecular markers are indicators of aggression and, most importantly, they can serve as guides for treatment.
In the past, preoperative systemic therapy was only used to treat locally advanced breast cancer, and patients with early-stage breast cancer (those with localized disease or disease that has spread only to regional lymph nodes) were candidates for breast-conserving surgery or mastectomy. More recently, preoperative systemic therapy, called neoadjuvant therapy, has become a common practice for patients with early-stage breast cancer. These treatments help to conserve breast and make inoperable tumors operable and are primarily recommended for triple-negative and HER2-positive breast cancer patients.
Clinical trials of neoadjuvant therapy for early breast cancer have also implemented the use of pathological complete responses (pCR). This indicates the absence of residual disease, an indicator of long-term outcomes and a clinical endpoint for drug approval.
In the United States, the National Comprehensive Cancer Network provides guidelines for therapeutic interventions for different subtypes of breast cancer. However, choosing between 12 different neoadjuvant regimens recommended for HER2-negative disease and 9 different neoadjuvant regimens recommended for HER2-positive disease demonstrates that clinicians are trying to choose the right treatment for each patient. face challenges.
Continuing approvals of new treatments and increasing standards of care make it imperative for clinicians to have better tools to guide and support decision-making so that patients are not over- or under-treated.
Predicting a patient’s response to different therapies provides clinicians with the opportunity to evaluate and compare different therapies. These types of techniques can guide clinicians when faced with critical decisions to escalate or downgrade treatment. This has important implications for prognosis and quality of life, as well as out-of-pocket costs incurred by patients.
Being able to predict the clinical outcomes of patients treated with standard-of-care therapies has value not only in the clinic, but also in clinical trials. For example, drug developers can use this feature to compare the benefits of a drug being tested to the current standard of care. They can use tools to predict the study group’s response to standard-of-care therapies to reveal the drug’s true benefit.
It can also be used in clinical trial design — predicting clinical outcomes — as a way to ensure that baseline responses to standards of care are similar across all study cohorts. In addition, choosing the right tools to predict patient outcomes can give stakeholders a more complete picture of the tumor that may go beyond outcome prediction.
Overall, it is important to have a deep understanding of how these methods work. This will improve the specificity and predictability of breast cancer treatment in clinical trials and in the clinic.
Image credit: andresr, Getty Images
