The use of magnetic resonance imaging (MRI) after neoadjuvant chemoimmunotherapy (NACI) may have significant benefit in forecasting treatment response in women with triple-negative breast cancer (TNBC)
For a new multicenter prospective study, recently published in Radiology, researchers assessed the prognostic capability of post-treatment MRI in 259 women who received NACI for TNBC. The study authors also assessed the predictive capacity of a model combining radiologic complete response (rCR) on post-treatment MRI, nodal status and the Ki-67 proliferation index. The total cohort was comprised of 175 women (mean age of 49) in the training set and 84 women (mean age of 52) in the training set, according to the study.
The study authors found that rCR, defined as no enhancement in the tumor bed, on breast MRI had a 83 percent AUC, 83 percent sensitivity, 84 percent specificity and a 90 percent positive predictive value (PPV) in predicting pathologic complete response (pCR) to NACI.
For the combined predictive model incorporating breast MRI rCR, nodal status and the Ki-67 proliferation index, researchers noted an 88 percent AUC, which was 23 percent higher than a radiopathologic prediction model (65 percent).
“ … Breast MRI after neoadjuvant chemoimmunotherapy shows high predictive value for pathologic complete response in patients with triple-negative breast cancer, particularly when combined with Ki-67 and nodal status,” wrote lead study author Toulsie Ramtohul, M.D., who is affiliated with the Department of Radiology at Institut Curie and PSL Research University in Paris, France, and colleagues.
Emphasizing that all of the included assessments in the combined predictive model were independently predictive for pCR, the researchers pointed out that multivariable analysis revealed that rCR on post-treatment breast MRI was 16.72 times more likely to be associated with pCR in patients with triple-negative breast cancer. The study authors noted a 4.69 times higher likelihood with a Ki-67 index above 30 percent, and a 2.48 times higher likelihood with the absence of nodal involvement.
The researchers also found a small false discovery rate (3.5 percent) among women who had rCR on breast MRI, no lymph node involvement and a Ki-67 index above 30 percent.
“… These patients could potentially avoid breast and axillary surgery after NACI, sparing (them) from breast deformity and unnecessary treatment-associated morbidity,” posited Ramtohul and colleagues.
Three Key Takeaways
1. High predictive accuracy of post-treatment MRI. Radiologic complete response (rCR) on post-NACI breast MRI demonstrated strong predictive performance for pathologic complete response (pCR), with 83 percent sensitivity, 84 percent specificity, and a 90 percent positive predictive value.
2. Enhanced prognostication with combined model. A predictive model combining rCR, nodal status, and Ki-67 index achieved an 88 percent AUC, which was 23 percent higher than a radiopathologic model. This highlights the added value of integrating imaging and biologic markers into predictive models.
3. Potential to avoid surgery in select patients. Women with rCR, no nodal involvement, and a Ki-67 index >30 percent had a very low false discovery rate (3.5 percent), suggesting the possibility of safely omitting breast and axillary surgery in exceptional treatment responders.
In an accompanying editorial, Natsuko Onishi, M.D., Ph.D., pointed out that no study participants with RCB-III tumors had rCR after NACI and the overall prognostic accuracy with post-treatment MRI demonstrated in the study may play a significant role in the management of patients with TNBC.
“Accurate prediction of pCR at breast MRI could significantly advance imaging-guided treatment decision-making, potentially leading to the elimination of breast surgery for exceptional responders after NAC and allowing for treatment redirection during neoadjuvant therapy. Achieving a high PPV while minimizing potential harm is key to realizing these paradigm shifts,” suggested Dr. Onishi, an imaging scientist in the Department of Radiology and Biomedical Imaging at the University of California, San Francisco (UCSF).
(Editor’s note: “Can Mid-Treatment MRI Help Predict Neoadjuvant Chemotherapy Response for Patients with Breast Cancer?,” “Study: Pre-Op MRI Makes No Difference in Outcomes for Women with HER-2 Positive, Hormone-Receptor Negative Breast Cancer” and “Can Abbreviated Breast MRI Have an Impact in Assessing Post-Neoadjuvant Chemotherapy Response?”)
In regard to study limitations, the study authors acknowledged modest sample sizes for subgroups with very low FDR, not including residual ductal carcinoma in situ in their definition of pathologic response, and varied scan acquisition parameters with different MRI scanner models.
