Prediction of pathological complete response after neoadjuvant chemotherapy in breast cancer by combining magnetic resonance imaging and core needle biopsy

BACKGROUND

Pathological complete response (pCR) is often achieved by neoadjuvant chemotherapy (NAC), particularly in hormone receptor-negative breast cancer. Contrast-enhanced magnetic resonance imaging (cMRI) is the most reliable imaging modality to evaluate the pathological effect of NAC. Ultrasonography is indispensable to collect representative specimens from the target lesion by core needle biopsy (CNB). This study aimed to evaluate the accuracy of predicting pCR by adding CNB after NAC, in cases with complete clinical response (cCR) diagnosed by cMRI.

METHODS

In this prospective multicentre study, we evaluated patients diagnosed with cCR by cMRI after NAC. Ultrasound-guided CNB (uCNB) using a 14G needle was performed without clip markers under general anaesthesia as planned surgery. Specimens collected by uCNB were compared to those resected surgically and were categorized as (i) no carcinoma (ypT0), (ii) no invasive carcinoma and only residual carcinoma in situ (ypTis) and (iii) residual invasive carcinoma. The concordance of pathological results between the uCNB and surgical specimens was evaluated.

RESULTS

Of the 83 patients evaluated, 41 (49.4%) and 17 (20.5%) of them had ypT0 and ypTis, respectively. The false negative rates (FNR), sensitivity and specificity for predicting ypT0 by uCNB were 50.0%, 50.0%, 100%, respectively, and those for predicting ypT0+ypTis were 28.0%, 72.0% and 98.3%, respectively. The concordance rates were 74.7% (62/83) for ypT0 and 90.4% (75/83) for ypT0+ypTis.

CONCLUSION

In cCR cases diagnosed by cMRI, uCNB was not accurate enough to predict pCR. Additional modalities like clip placements and/or thicker core needles may be required for better prediction.