Background: Adjuvant polychemotherapy (ACT) decreases the risk for recurrence and improves survival of patients operated for breast cancer. While the relative risk reduction is similar regardless of any clinicopathologic factor, the absolute benefit is a function of both the absolute risk and the expected toxicities. Among the chemotherapy induced toxicities, the most common side effect is myelotoxicity which can result in anemia, thrombopenia and leukopenia. Severe neutropenic or leukopenic events (NLEs) may be complicated by life-threatening infections (febrile neutropenia, FN), requiring hospitalization and antibiotic therapy. The risk for FN and NLEs is assessed using clinical risk factors, and risk prediction is included in the current guidelines. Interestingly, myelotoxicity in general and neutropenia, particularly during ACT have been shown to be predictors for improved outcomes. This association between the toxicity and survival has been demonstrated in several retrospective studies and supports the use of hematologic suppression as a tool for personalized dosing of ACT. The feasibility of this approach was initially assessed in the randomized phase 2 SBG 2004-1 trial and its efficacy in the phase 3 PANTHER trial. In PANTHER, a standard schedule of adjuvant chemotherapy (FEC x 3 followed by Docetaxel x 3, every three weeks) was compared with an investigational dose dense and tailored schedule (EC x 4 followed by Docetaxel x 4 every two weeks). In the investigational arm, a tailored dosing was made based on bone-marrow toxicity seen after each cycle. Our Hypothesis is that genetic variation can predict chemotherapy-induced hematological toxicity. Genetic biomarkers can also help us to understand the mechanisms underlying chemo-induced hematological toxicity. We aim to perform association studies of the SNPs with the phenotypes of interest (grade 3 and 4 neutropenia and leukopenia) in the phase 2 SBG 2004-1 and the phase 3 PANTHER cohorts to validate top significant SNPs from previous studies, but also to identify additional SNPs. As an exploratory secondary endpoint, we also aim to assess the predictive value of selected SNPs for tailored and dose-dense chemotherapy in PANTHER study. Blood samples from trial participants have been collected to our laboratory at Karolinska and germline DNA has been extracted and analyzed with genotyping using the Global screening Array- Multiple Disease (GSA-MD) from Illumina (730K SNPs, customized to include all SNPs from preliminary data). Clinical follow-up in the trials is available.