591 Introduction: Historically, pCR is rare in AC-Res in contrast to AC-S BC except following TCH in HER2+ BC in preliminary analysis (SABCS-2004, #1110). Moreover, robust predictors of pCR are needed. Methods: 106 consecutive BC pts on neoadjuvant studies were treated with GM/G-CSF supported dose-dense AC (except 7**- Table ), 2 cycles if AC-Res and 4 if AC-S. Pts then received 9–12 wkly TC (3 wks on, 1 off) ± 6–8 bi-wkly B if HER2- or +12–16 wkly H if HER2+, followed by surgery. Fisher’s exact test was performed to compare pCR percentages by various characteristics. Results: pCR in breast and lymph nodes in 38 of 84 assessable tumors in the first 82 pts were documented. 67% (56/84) tumors were reduced to ≤5 mm. Overall, no difference in pCR rates were found between AC-S and AC-Res BC, but pCR rates were 2- to 4-fold higher in HER2+, AC-S (79%) and AC-Res (65%) subsets compared to HER2-, AC-S (32%) and AC-Res (16%) subsets. Higher pCR rates were associated with HER2+/hormone receptor- (HR-) > HER2+/HR+, HER2-/HR- (triple-) > HER2-/HR+, and within the HER2+ subset-IHC 3+/FISH+ or unknown > IHC 3+/FISH- or IHC 1–2+/FISH+, and ductal > other histology. No patient had clinical cardiac dysfunction or EF <50, except 1 on AC/TC. 13/106 (12%) progression or death at a median follow-up of 22 months (range 2–46) is mostly due to stage IV and CNS progression. Conclusion: Short course of TCH achieves a high rate of pCR in AC-Res and AC-S, HER2+ BC in contrast to TC ± B in AC-Res and AC-S, HER2- BC. This is the first demonstration of high rate of complete AC resistance reversal across stage II-IV, inflammatory and recurrent HER2+ BC. In vivo response adjusted 2–4 cycles of dose-dense AC limited clinical cardiac toxicity. HER2+ (IHC 3+/FISH+ or unknown), HR- and ductal histology are significant predictors of high pCR. HER2 and HR confer 4 predictive subtypes of BC. [Table: see text] No significant financial relationships to disclose.