Abstract 219P
Background
The presence of radiologic complete response (rCR) on magnetic resonance imaging (MRI) after neoadjuvant chemotherapy (NAC) may help to identify breast cancer patients in whom surgery can be safely omitted. This systematic review and meta-analysis evaluate the accuracy of MRI for pathological complete response (pCR) after NAC in breast cancer immunohistochemical (IHC) subtypes, which represent distinct phenotypes with different responses to NAC.
Methods
Two researchers systematically searched PUBMED and EMBASE to select relevant studies and extract data. We categorized study results based on hormone receptor (HR) and HER2-receptor status. For meta-analysis of sensitivity and specificity according to IHC subtype, we used bivariate random-effects models.
Results
26 studies (6151 patients) were included. The accuracy of MRI for post-NAC pCR depended on IHC subtype (likelihood ratio p=0.0082; Table). Table: 219P
Summary effect measures of the bivariate random-effects model
| Subtype | Pooled Sensitivity for detecting pCR [95% CI] | Pooled Specificity for detecting pCR [95% CI] |
| HR-HER2-1509 patients/21 studies | 0.66 [0.55-0.75] | 0.84 [0.80-0.88] |
| HR- HER2+855 patients/14 studies | 0.65 [0.54-0.74] | 0.81 [0.74-0.87] |
| HR+ HER2-1988 patients/14 studies | 0.52 [0.42-0.62] | 0.88 [0.84-0.91] |
| HR+ HER2+902 patients/12 studies | 0.64 [0.50-0.76] | 0.71 [0.57-0.82] |
Specificity was significantly lower in the HR+HER2+ subtype compared to the HR-HER2- (p=0.0012) and HR+HER2- subtype (p<0.001), and borderline significantly lower compared to the HR-HER2+ subtype (p=0.059). Sensitivity was overall moderate and lowest in the HR+HER2- subtype. There was no significant difference in sensitivity between the different IHC subtypes. Based on these findings and published pCR rates, the positive predictive value of MRI for pCR is projected to be 68% in the HR-HER2-, 78% in the HR-HER2+, 37% in the HR+HER2-, and 50% in the HR+ HER2+ subtype.
Conclusions
The proportion of patients in whom surgery would be justly omitted based on rCR on MRI after NAC is lowest in the HR+HER2- and highest in the HR-HER2+ subtype, but overall too low to safely omit surgery after NAC based on MRI alone. Implementation of additional non-invasive response evaluation techniques, such as advanced image analysis and/or liquid biopsies, is needed.
Clinical trial identification
Editorial acknowledgement
Legal entity responsible for the study
The authors.
Funding
European Union Horizon 2020 research and innovation program under grant agreement no. 755333.
Disclosure
All authors have declared no conflicts of interest.