New Approach May Help Predict HER2+ Tumor Response to Treatment

A newly identified radiogenomic signature from human epidermal growth factor receptor 2 -positive (HER2+) breast cancer tumors and their surrounding environment could serve as a future noninvasive method for predicting response to targeted treatment.

The emergence of HER2-targeted therapy, including the monoclonal antibodies trastuzumab (Herceptin) and pertuzumab (Perjeta), has greatly improved survival in HER2+ breast cancer. Yet, a significant percentage of patients will not achieve a complete preoperative response to a combination of anti-HER2 therapy and chemotherapy, and no clinically validated biomarker is currently available to indicate which patients are likely to benefit from targeted therapies.

Breast radiogenomics — an investigational diagnostic approach that integrates genomic data and qualitative analysis of clinical radiology for tumor characterization — has shown promise in noninvasively identifying patients’ genetic profile from imaging, but has not been applied in the context of predicting clinical outcomes and guiding targeted therapies.

Now, a multicenter team including researchers from Cleveland Clinic has shown that a combination of measurements within and outside a tumor on clinical dynamic contrast-enhanced magnetic resonance imaging (DCE–MRI) is capable of discriminating the response-associated HER2-enriched molecular subtype from other subtypes among patients with HER2+ tumors. (Current subtype identification requires costly gene expression profiling using tissue obtained by an invasive biopsy.)

When subsequently evaluated among recipients of HER2-targeted therapy, the new intratumoral and peritumoral imaging signature was found to be associated with response to neoadjuvant chemotherapy.

“Currently, if we see someone with a HER2-positive tumor, we always just give them chemotherapy and HER2-targeted medicine,” says Jame Abraham, MD, Director of Cleveland Clinic Cancer Center’s Breast Oncology Program and Co-Director of the Comprehensive Breast Cancer Program. “Until now, no one has looked at a predictive model to see who will benefit. This is the first use of radiology and radioanalysis to identify that subset of patients.”

Potential treatment impacts

The approach uses computerized tissue phenotyping on radiographic imaging (radiomic) features extracted from breast MRI to examine the appearance of the tumor and its surroundings. A combination of local disorder, especially within the peritumoral environment, and larger-scale homogeneity near the tumor were found to most effectively characterize the treatment response-associated HER2-enriched (HER2-E) molecular subtype.

Co-occurrence of local anisotropy gradients (CoLlAGe) feature expression maps visualize the elevated disorder of local intensity gradient orientations within the peritumoral region of HER2-E relative to non–HER2-E breast cancers. Radiomic feature values are unitless, thus the scale depicts relative expression values of radiomic features, standardized between 0 and 1.0 based on the range of their distribution. The blue color at 0 depicts the minimum observed feature value; the red color at 1.0 depicts the maximum observed feature value.
Credit: Braman N, Prasanna P, Whitney J, et al. Association of Peritumoral Radiomics With Tumor Biology and Pathologic Response to Preoperative Targeted Therapy for HER2 (ERBB2)–Positive Breast Cancer. JAMA Netw Open. 2019 Apr 5;2(4):e192561.

If confirmed in subsequent studies, the technique could alter the course of treatment, since patients identified with non-responsive HER2 molecular subtypes could be sent for surgical resection first, and other treatments afterward, notes Abraham, who is also Professor of Medicine at Cleveland Clinic Lerner College of Medicine.

Alternatively, this approach could be used to identify patients who might benefit from trastuzumab emtansine (T-DM1), an antibody–drug conjugate of trastuzumab and the cytotoxic agent emtansine (DM1), a maytansine derivative and microtubule inhibitor. In a landmark study published in December 2018, T-DMI reduced the risk of recurrence of invasive breast cancer or death by 50% in patients with HER2+ early breast cancer who had residual invasive disease after completion of neoadjuvant therapy compared to trastuzumab alone.

“We could potentially select patients up front to treat with T-DMI. This would represent a major step for personalized medicine,” Dr. Abraham says.

Identifying Responders Radiographically

Initially, the investigators identified imaging features distinguishing HER2+ tumors from other receptor subtypes among 117 patients who received an MRI prior to neoadjuvant chemotherapy at a single institution between 2012 and 2015.

Imaging signature of HER2-E is associated with pathologic complete response (pCR) to anti-
HER2 therapy, with rippled enhancement patterns detected intratumorally by Laws feature (left), and elevated local peritumoral heterogeneity captured by CoLlAGe features 9 to 12 mm from the tumor (right) characterizing both features. Radiomic feature values are unitless, thus the scale depicts relative expression values of radiomic features, standardized between 0 and 1.0 based on the range of their distribution. The blue color at 0 depicts the minimum observed feature value; the red color at 1.0 depicts the maximum observed feature value.
Credit: Braman N, Prasanna P, Whitney J, et al. Association of Peritumoral Radiomics With Tumor Biology and Pathologic Response to Preoperative Targeted Therapy for HER2 (ERBB2)–Positive Breast Cancer. JAMA Netw Open. 2019 Apr 5;2(4):e192561.

Then, using imaging and genomic data from a previous multicenter trial of 42 patients with HER2+ breast cancer and preoperative MRI and RNA sequencing data, they developed a signature to identify the HER2-E subtype among clinically HER2+ tumors. Previous radiomics studies have focused on analyzing the tumor itself, but the team found that adding information about its surroundings was critical to distinguishing HER2-E tumors.

To evaluate the utility of this signature in guiding treatment decisions, the team explored whether it could be used to predict targeted therapy outcome for HER2+ patients. When applied to a set of 78 patients from two institutions who had received MRI exams before HER2-targeted neoadjuvant chemotherapy, the signature was found to significantly identify patients who would achieve a complete response.

To better understand this signature, the researchers compared radiomic features with biopsy samples from the same patients. They observed that features from the 0-3mm peritumoral region on MRI were significantly associated with the density of tumor-infiltrating lymphocytes on tissue samples — indicating a potential link between the way the immune system responds to a tumor and the appearance of its surroundings on imaging.

While the findings are compelling, Dr. Abraham cautions that “This is a completely experimental, retrospective study — a proof of concept. We need to validate this in larger datasets and confirm the data. Then we can potentially use it for wide application.”