Adjuvant chemotherapy for breast cancer – prognostic impact of relative dose-intensity.

Abstract

Abstract Abstract #4109 Introduction: Maintenance of adequate dose intensity (DI) is an important aim in the use of adjuvant chemotherapy for early breast cancer. Some studies report a decreased survival, if the DI is below 80% of the planned dose. We evaluated the rate of decreased DI and its prognostic impact for patients (pts.) treated in our institution between 1990 and 2004. During the long observation period a variety of regimens were used: CMF, normal dosed anthracycline (norA) based regimens (Epirubicin less than 35 mg/sqm*week) and higher dosed anthracyclines (hiA; Epirubicin doses of more than 45 mg/sqm*week).
 Methods: All patient files were reviewed to calculate relative DI; cycles not applied were treated as a 100% dose reduction. Follow up-data were taken from our institutional prospective data base and the regional tumor registry. Excluded from the analysis were pts. with high dose chemotherapy with stem cell support, rare regimens (less than 10 cases / regimen), pts. switching institution during therapy and pts. progressing during chemotherapy.
 Results: From 966 pts. receiving adjuvant chemotherapy, 887 matched the inclusion criteria. Mean follow up is 70 months with 260 distant progressions and 234 deaths. 193 pts. received CMF, 442 norA and 252 hiA, 302 pts. received taxanes. Preterm termination of chemotherapy occurred in 77 pts (8.7%), dose reductions in 36 pts. (4.1%). Pts. receiving hiA were at higher risk for dose reductions than with norA and CMF (7.5% vs. 2.9% vs. 2.1%, p=0.015). Mean dose delays were 12.7 days (CMF), 6.4 days (norA) and 5.7 (hiA) (p<0.01, Kruskal-Wallis-test). The mean rel. DI were 90% (CMF), 93% (norA) and 92% (hiA) and significantly different (p=0.01, Kruskal-Wallis-test).
 In univariate analysis, 5 year-metastasis free survival (MFS, 74% vs. 55%, p=0.002) and overall survival (OAS, 80% vs. 65%, p=0.009) are significantly lower for pts. with a rel. DI of less than 80%. This effect seems more pronounced for norA (OAS 80% vs. 45%), than for CMF (OAS 74% vs. 74%) and hiA (OAS 85% vs. 77%). The adverse effect of low DI was independent of taxan-use. In a cox-regression analysis low DI was not identified as independent predictor for MFS and OAS.
 Conclusions: Compared to DI reported from most prospective randomised trials the DI in this study seems lower, potentially reflecting patient selection. Dose reductions, delays and preterm termination are common findings, leading to impaired DI. In our study low DI is associated with adverse prognosis in univariate, but not in multivariate analysis. Nevertheless, unnecessary dose modifications should be avoided. Citation Information: Cancer Res 2009;69(2 Suppl):Abstract nr 4109.