Abstract P3-12-34: Antibiotic use and mortality from triple-negative breast cancer

Abstract

Abstract Background: Gut-associated lymphoid tissue is the largest component of the body's immune system, as it influences both local and systemic immune responses. Gut microbiome dysbiosis related to antimicrobial exposure may be associated with decreased circulating and tumor-infiltrating lymphocytes and decreased immune repertoire, which may adversely impact survival in patients with triple-negative breast cancer (TNBC). We hypothesized that increasing antimicrobial exposure may lead to higher overall mortality (OM) and BC-specific mortality (BCM) in the presence of time-varying absolute lymphocyte count (ALC). Methods: Women with TNBC were identified in the Oncoshare database, a breast cancer registry that integrates data from the population-based SEER Registry and electronic medical records from two California healthcare systems (Stanford University and Sutter Health). We defined antibiotic use in three ways at each time point: (1) current use (“Antibiotic Use”), (2) total number of prescriptions (“Total Antibiotics”), and (3) total number of unique antibiotics prescribed (“Unique Antibiotics”) to study overall OM and BCM. We used marginal structural Cox proportional hazards multivariate models, with time-varying covariates (antibiotic use and ALC) to avoid immortal time bias. Results: 772 women were diagnosed with TNBC and treated from 2000-2014. The median overall follow-up time (including time to death) was 104 (IQR [61.66, 147.03]) months; the median follow-up time among those who were alive through the observation period was 121 (IQR [86.89, 160.93]) months. There were 654 (85%) patients who ever used antibiotics after TNBC diagnosis. There were 24/118 (20%) deaths among patients who never took antibiotics and 153/654 (23%) deaths in patients who ever took antibiotics during the study period. We previously showed that higher minimum ALC was associated with lower OM (HR 0.23, 95%CI (0.16, 0.35)) and BCM (HR 0.19, 95%CI (0.11, 0.34)) in TNBC patients from this cohort, with a significant association between higher peripheral lymphocyte count and tumor-infiltrating lymphocytes. Here, we found antibiotic use was associated with higher OM and BCM using definitions (2) and (3), but not definition (1). Model results are shown in the table and individual covariates are included for all definitions of antibiotic use. Conclusion: Higher number of antibiotic prescriptions and of unique antibiotics prescribed was associated with overall and breast cancer-specific mortality among women with TNBC. Future research on the role of the microbiome in mediating ALC and immune response may inform interventions to reduce TNBC mortality. Table. Overall and breast cancer-specific mortality among women with TNBC, Cox proportional hazards modelsOverall MortalityAntibiotic UseTotal AntibioticsUnique AntibioticsAntibiotic use1.54 (0.99, 2.39)Total number of antibiotics1.07* (1.04, 1.09)Number of unique antibiotics1.17* (1.12, 1.23)Age at diagnosis1.01* (1.00, 1.03)†1.01 (1.00, 1.02)1.01 (1.00, 1.03)Race/ethnicity (reference=White)Hispanic1.25 (0.70, 2.24)1.26 (0.70, 2.26)1.23 (0.69, 2.22)Non-Hispanic Asian Pacific Islander1.01 (0.64, 1.61)1.16 (0.74, 1.80)1.10 (0.69, 1.73)Non-Hispanic Black2.07* (1.12, 3.82)2.01* (1.10, 3.68)1.87* (1.03, 3.38)Neighborhood socioeconomic status quintile (reference=lowest)Second-lowest1.15 (0.49, 2.73)1.08 (0.47, 2.51)1.10 (0.49, 2.47)Middle1.22 (0.53, 2.81)1.05 (0.46, 2.40)1.06 (0.48, 2.35)Second-highest0.78 (0.34, 1.83)0.69 (0.30, 1.59)0.70 (0.31, 1.56)Highest0.96 (0.44, 2.11)0.82 (0.37, 1.78)0.80 (0.38, 1.70)Stage (reference=I)II1.72* (1.14, 2.59)1.61* (1.05, 2.45)1.71* (1.13, 2.57)III4.93* (3.08, 7.89)4.99* (3.10, 8.04)5.15* (3.20, 8.29)Grade (reference= 1)21.84 (0.66, 5.12)1.70 (0.61, 4.73)1.61 (0.57, 4.59)32.56 (0.95, 6.89)2.37 (0.88, 6.38)2.24 (0.82, 6.13)Received chemotherapy0.72 (0.45, 1.17)0.61 (0.37, 0.99)0.66 (0.41, 1.08)Received radiotherapy0.97 (0.70, 1.35)0.96 (0.69, 1.35)0.99 (0.71, 1.36)Used growth factors (e.g., filgrastim)1.38 (0.97, 1.97)1.31 (0.90, 1.93)1.15 (0.79, 1.67)† 95% CI lower bound is rounded from 1.001. * p < 0.05 Breast Cancer-Specific MortalityAntibiotic UseTotal AntibioticsUnique AntibioticsAntibiotic use1.50 (0.90, 2.47)Total number of antibiotics1.07* (1.04, 1.10)Number of unique antibiotics1.18* (1.12, 1.25)Age at diagnosis1.00 (0.98, 1.01)1.00 (0.98, 1.01)1.00 (0.98, 1.01)Race/ethnicity (reference=White)Hispanic1.36 (0.71, 2.61)1.36 (0.71, 2.60)1.34 (0.69, 2.58)Non-Hispanic Asian Pacific Islander1.41 (0.86, 2.31)1.48 (0.91, 2.40)1.47 (0.90, 2.40)Non-Hispanic Black1.69 (0.76, 3.76)1.66 (0.76, 3.60)1.55 (0.70, 3.43)Neighborhood socioeconomic status quintile (reference=lowest)Second-lowest1.54 (0.52, 4.52)1.71 (0.57, 5.17)1.59 (0.53, 4.79)Middle1.50 (0.53, 4.26)1.61 (0.55, 4.71)1.48 (0.51, 4.31)Second-highest1.15 (0.41, 3.27)1.17 (0.40, 3.41)1.12 (0.39, 3.26)Highest1.14 (0.42, 3.06)1.19 (0.43, 3.30)1.05 (0.38, 2.92)Stage (reference=I)II2.39* (1.37, 4.17)2.43* (1.39, 4.27)2.45* (1.40, 4.28)III7.96* (4.31, 14.70)8.35* (4.49, 15.53)8.48* (4.55, 15.80)Grade (reference= 1)21.12 (0.37, 3.36)1.01 (0.33, 3.05)0.91 (0.30, 2.81)31.67 (0.60, 4.68)1.52 (0.54, 4.26)1.42 (0.50, 4.04)Received chemotherapy0.64 (0.36, 1.14)0.61 (0.34, 1.10)0.60 (0.33, 1.09)Received radiotherapy0.95 (0.64, 1.39)0.91 (0.62, 1.33)0.95 (0.65, 1.38)Used growth factors (e.g., filgrastim)1.05 (0.70, 1.57)0.94 (0.63, 1.41)0.88 (0.58, 1.33)NOTE: * denotes p<0.05 Citation Format: Julia D Ransohoff, Karen Andrade, Natasha Purington, Vidhya Balasubramanian, Summer Han, Mina Liu, Jennifer Caswell-Jin, George Sledge, Ami Bhatt, Allison Kurian. Antibiotic use and mortality from triple-negative breast cancer [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr P3-12-34.