Return To Abstract Listing

Scientific Session 17 — Breast -MRI/New Technologies

Wednesday, May 3, 2017

Abstracts 1031-2861

2511. Prognostic Role of Detection Method in Invasive Breast Cancer

James L*,  Lewis M,  Leddy R,  Collins H,  Ackerman S. Medical University of South Carolina, Charleston, SC

Address correspondence to M. Lewis (

Objective: The purpose of this study was to assess the impact of tumor detection method in predicting the need for axillary lymph node dissection (ALND) and chemotherapy.

Materials and Methods: A total of 263 women (age range, 40–87 years) with invasive breast cancer were identified through our local cancer registry database from July 1, 2014, to June 30, 2015. Deidentified patient data included age at diagnosis, race, method of cancer identification, derived American Joint Committee on Cancer stage, biomarkers, chemotherapy administration, and ALND. A chi-square analysis was used to test the strength of association between ALND by detection method and chemotherapy administration by detection method. Statistical significance was considered at the α = 0.05 threshold, and two-sided exact p values are reported. Effect size was evaluated with Cramer V and are interpreted by Cohen’s standard guidelines of small (0.1), medium (0.3), and large (0.5). All statistical analyses were conducted with SPSS version 23 (IBM).

Results: In the 263 women with invasive breast cancer, 58% (n = 152) of cancers were screen-detected and 42% (n = 111) were detected by other methods. There was a significant association between ALND and detection method (χ2 [df, 1, N = 263] = 21.69, p < 0.001). Of the patients undergoing ALND, 37.84% (nn = 42) had symptomatic breast cancers compared with only 13.14% (n = 20) of screening-detected patients. This is a small to medium effect of detection method on ALND (Cramer V = 0.29). There was a significant association between chemotherapy and detection method (χ2 [df, 1, N = 263] = 32.11, p < 0.001). Of the patients receiving chemotherapy, 60.36% (n = 67) of patients with symptomatic cancers received chemotherapy compared with only 25.66% (n = 39) of screening-detected patients. This is a medium effect of detection method (Cramer V = 0.35). In addition, logistic regression analyses were conducted to identify the effects of detection method, age, stage, and biomarkers on predicting ALND or chemotherapy. Overall, the models were statistically significant and correctly classified 84.4% of axillary dissection cases and 84.8% of chemotherapy cases. However, only stage was found to significantly contribute to the model for predicting ALND, whereas stage, age, and biomarkers significantly contributed to the predictive model for receiving chemotherapy.

Conclusion: Screening-detected invasive breast cancers are less likely to undergo ALND and receive chemotherapy compared with cancers presenting symptomatically, which decreases morbidity associated with their treatment course.