Background: Patients achieving a pathologic complete response (pCR) followingneoadjuvant therapy have significantly improved event-free survival relative tothose who do not; and pCR is an FDA-accepted endpoint to support acceleratedapproval of novel agents/combinations in the neoadjuvant treatment of high riskearly stage breast cancer.  Previous studieshave shown that recurrence risk increased with increasing burden of residualdisease (as assessed by the RCB index). As well, these studies suggest that patients with minimum residualdisease (RCB-I class) also have favorable outcomes (comparable to thoseachieving a pCR) within high risk tumor subtypes. In this study, we assesswhether integrating RCB with MRI functional tumor volume (FTV), which in itselfis prognostic, can improve prediction of distant recurrence free survival(DRFS); and identify a subset of patients with minimal residual disease withcomparable DRFS as those who achieved a pCR. Imaging tools can then be used toidentify the subset that will do well early and guide the timing of surgicaltherapy.

Method: We performed a pooled analysis of 596 patientsfrom the I-SPY2 TRIAL with RCB, pre-surgical MRI FTV data and known follow-up (median2.5 years). We first assessed whether FTV predicts residual disease (pCR orpCR/RCB-I) using ROC analysis. We applied a power transformation to normalizethe pre-surgical FTV distribution; and assessed its association with DRFS usinga bi-variate Cox proportional hazard model adjusting for HR/HER2 subtype. Wealso fitted a bivariate Cox model of RCB index adjusting for subtype; andassessed whether adding pre-surgical FTV to this model further improvesassociation with DRFS using a likelihood ratio (LR) test.  For the Cox modeling, penalized splinesapproximation of the transformed FTV and RCB index with 2 degrees of freedomwas used to allow for non-linear effects of FTV and RCB on DRFS.  

Result: Pre-surgical MRI FTV is significantlyassociated with DRFS (Wald p<0.00001), and more effective at predicting pCR/RCB-Ithan predicting pCR alone (AUC: 0.72 vs. 0.65). Larger pre-surgical FTV remains associated with worse DRFS adjusting forsubtype (Wald p <0.00001).  The RCBindex is also significantly associated with DRFS adjusting for subtype (Wald p<0.00001).Adding FTV to a model containing RCB and subtype further improves associationwith DRFS (LR p=0.0007).  RCB-I patientshave excellent DRFS (94% at 3 years compared to 95% in the pCR group).  Efforts are underway to identify an optimalthreshold for dichotomizing pre-surgical FTV and FTV change measures for use incombination with pCR/RCB-I class to generate integrated RCB (iRCB) groups as acomposite predictor of DRFS.

Conclusion: Pre-surgical MRI FTVis effective at predicting minimal residual disease (RCB0/I) in the I-SPY 2TRIAL.  Despite the association betweenFTV and RCB, FTV appears to provide independent added prognostic value (to RCBand subtype), suggesting that integrating MRI volume measures and RCB into acomposite predictor may improve DRFS prediction.