Multiple Factors Predict Recurrence Following Neoadjuvant Chemotherapy

The combination of multiple independent factors minimized the impact of baseline stage for predicting locoregional recurrence (LRR) following neoadjuvant chemotherapy in patients with breast cancer.

Eleftherios P. Mamounas

The combination of multiple independent factors minimized the impact of baseline stage for predicting locoregional recurrence (LRR) following neoadjuvant chemotherapy in patients with breast cancer, according to a large retrospective analysis presented in advance of the 2014 Breast Cancer Symposium.

In the study, overall rates of LRR at 5 years were below 10% for patients treated with neoadjuvant chemotherapy, with breast cancer subtype and pathologic complete response (pCR) representing the most predictive independent factors for recurrence. Based on type of surgery, age was shown to be a predictor of LRR for patients treated with lumpectomy but not mastectomy. Overall, the inclusion of pCR and subtype minimized the predictive qualities of baseline stage, Eleftherios Mamounas, MD, MPH, explained during a presentation of the results.

Learn more about recurrence and tailoring radiotherapy following neoadjuvant chemotherapy

"After we took into account the tumor subtypes, the types of surgery, and the pathologic response status, the stage of presentation before neoadjuvant therapy was no longer an independent predictor," Mamounas, the medical director of the Comprehensive Breast Program at the University of Florida Health Cancer Center, said. "Our findings have clinical implications relative to further tailoring the use of adjuvant radiotherapy after neoadjuvant chemotherapy and further support the conduct of ongoing clinical trials that attempt to tailor locoregional therapy in the neoadjuvant chemotherapy setting."

The analysis looked at data for 11,955 patients across 12 large neoadjuvant breast cancer trials. The median follow-up for patients in the study was 5.4 years. The median age of patients was 49 years, with 61% of patients having T2 tumors and 47% having clinically node-positive disease.

In a 5252 patient 5-year cumulative analysis, LRR rates were 8.8%. LRR occurred in 7.8% of patients treated with lumpectomy and 10.4% with mastectomy. Age, tumor subtype, baseline stage, and pCR status all represented independent predictors of LRR by multivariate analysis.

"Obviously these two groups of patients are not comparable, since usually patients with more advanced disease end up having mastectomy," Mamounas explained.

Patients with the HER2-positive, HR-negative (n = 709) subtype of breast cancer were most likely to experience LRR, at 14.8%. Additionally, patients with triple-negative breast cancer (HR-/HER2-negative) had a LRR rate of 12.2%. Overall, for lumpectomy, the strongest predictor of recurrence was the HR-negative, HER2-positive subtype, with a hazard ratio (HR) of 7.90 (P<.0001). In the mastectomy arm, pCR status was the most predictive factor, with an HR of 3.88 (P<.0001).

"The patients with HER2-positive breast cancer in this analysis, about a third of these patients or so received adjuvant trastuzumab, so these HER2-positive patients were not treated as they would be treated today," Mamounas said. “While more research is needed to inform new practice guidelines based on these insights, the findings provide additional information for doctors and patients to consider when trying to decide on the best locoregional treatment options after neoadjuvant chemotherapy.”

The utility of pCR for predicting outcomes to neoadjuvant therapy was recently called into question by discordant results from the phase III ALTTO and NeoALTTO studies. In NeoALTTO, neoadjuvant lapatinib combined with trastuzumab demonstrated a marked improvement in pCR and produced significantly higher 3-year event-free survival rates, suggesting similar activity in the adjuvant setting. However, the combination of lapatinib and trastuzumab failed to demonstrate a significant improvement in disease-free survival compared with trastuzumab alone.

"Whether a patient that has a pCR does better than a patient who does not have a pCR is absolutely true in the majority of subtypes of breast cancer," Mamounas said. "When it comes to HER2-positive patients, triple-negative patients, or high-proliferative ER-positive patients, clearly those who achieve a pCR do significantly better than those who not, that much is clear from this analysis and any other study that has been published on that subject."

Mamounas EP, Cortazar P, Zhang L, et al. Locoregional recurrence (LRR) after neoadjuvant chemotherapy (NAC): Pooled-analysis results from the collaborative trials in neoadjuvant breast cancer (CTNeoBC). Presented at: 2014 Breast Cancer Symposium; September 4 - 6, 2014; San Francisco, CA. Abstract 61.

<<<Back to the 2014 Breast Cancer Symposium conference page