Combining a liquid biopsy with magnetic resonance imaging (MRI) significantly improved accuracy for predicting complete response among patients with locally advanced breast cancer who received neoadjuvant chemotherapy, according to a study presented during the AACR Annual Meeting 2021.
The improved accuracy may spare patients of a sentinel lymph nodes biopsy in the armpit, performed to confirm cancer has not spread to the axillary lymph nodes, the researchers said.
"Radiological procedures such as MRI can be potentially combined with the assessment of circulating biomarkers in order to advance overall accuracy," said Francesco Ravera, MD, PhD, a fellow in the Department of Internal Medicine at the University of Genoa in Italy. "This can bring effective advancements in the pursuit of a personalized and minimally invasive medicine."
Current standard of care to assess a breast cancer patient's response to chemotherapy prior to surgery, or neoadjuvant chemotherapy, includes radiological imaging of the tumor site, such as MRI. The procedure, designed to detect and measure tumor size and location, has an overall accuracy of about 80 percent in correctly assessing pathological complete response, defined as the absence of invasive cancer in breast and axillary nodes.
"Despite being the most accurate radiological procedure for this purpose, MRI does not allow to spare sentinel lymph node biopsy in patients assessed as complete responders, possibly exposing patients to collateral side effects and relevant discomfort," Ravera said in an interview.
Previous studies have demonstrated the potential efficacy of liquid biopsies, particularly plasma cell-free DNA (cfDNA) integrity, as a biomarker for the early diagnosis, recurrence, and response to therapy. Low cfDNA integrity, which corresponds to high cfDNA fragmentation, is a typical feature of neoplastic patients. When healthy cells die, they typically release similarly sized DNA fragments into the bloodstream. However, when cancer cells die, they release DNA fragments of varying sizes. By measuring the quantity of different fragmented sizes, clinicians can estimate the integrity of patients' cfDNA.
To estimate the accuracy of cfDNA integrity in assessing response of neoadjuvant therapy among breast cancer patients, Ravera and colleagues evaluated plasma taken from 38 patients who had completed an anthracycline/taxane-based treatment prior to surgery. Concentrations of different cfDNA fragments were assessed, with fragments ranging in size from 90-150 base pairs (bp), 150-220 bp, 221-320 bp, 100-300 bp, and 100-300 bp.
A score known as a cfDNA integrity index (DII) is typically expressed as the ratio of the concentration of long cDNA fragments to short cfDNA fragments. Fragment size ranges evaluated as the most informative indicators of response to neoadjuvant therapy were selected to calculate a measure of DNA integrity, expressed as a ratio of 321-1,000 bp concentration to 150-220 bp size fragments.
"The concentrations of 150-220 bp size fragments, typically derived from cellular apoptosis, significantly varied from pathological complete response patients and was therefore selected as a parameter to calculate cfDNA integrity index," Ravera said. "Fragments ranging from 321 to 1,000 bp presumably derived instead from non-apoptotic processes of cellular death.
"The cfDNA integrity index was therefore calculated as the ratio of the concentration of presumably necrotic cfDNA fragments to the one of fragments derived from cellular apoptosis."
Among the 38 patients evaluated, 11 experienced pathological responses following neoadjuvant chemotherapy, while 27 patients experienced an incomplete response, with residual disease either in the breast or axillary nodes following treatment. MRI had an accuracy of 77.1 percent, while the cfDNA integrity index of DII had an accuracy of 81.6 percent in predicting the achievement of a complete response at histopathological examination.
Ravera and colleagues then evaluated whether the cfDNA integrity index could be combined with MRI to improve prediction. The two concordant techniques yielded a combined prediction of complete response achieved with an accuracy of 92.6 percent, with a positive predictive value (accuracy in predicting a positive result) and a negative predictive value (accuracy in predicting a negative result) of 87.5 percent and 94.7 percent, respectively.
"Our work identifies a new parameter that is easily combinable with MRI to a more accurate prediction of response following neoadjuvant treatment, with possible implications for current protocols for the evaluation of nodal residual disease among patients with breast cancer undergoing neoadjuvant chemotherapy," Ravera noted.
Next steps include the confirmation of these results after the expansion of sample sizes, including evaluation of its use in prospective studies.
"In case of effective confirmation, the evaluation of the effective clinical validity of the combination of cfDNA integrity index and MRI would be of great interest," Ravera said. "This may be assessed by sparing sentinel lymph node biopsy in those patients assessed as complete responders by both techniques, comparing the rate of local recurrence with patients who undergo the usual clinical procedures."
Warren Froelich is a contributing writer.