408 Functional Status and Associated Treatment Patterns ...

1 408 Functional Status and Associated Treatment Patterns among metastatic Triple Negative Breast Cancer in EU 5. Della Varghese1; Kala Hill2; Marc Botteman1. 1. Pharmerit International, Bethesda, MD, USA. 2. Celldex Therapeutics, Hampton, NJ, USA. BACKGROUND RESULTS RESULTS (CONTINUED) RESULTS (CONTINUED) RESULTS (CONTINUED). Triple negative breast cancer (TNBC) is characterized by negative Patient Characteristics Among ECOG 0-1 patients, the most common treatments were Logistic regression analyses showed: estrogen, progesterone, and HER2 receptors. It affects 15% of Table 2: Physician Characteristics (Figure 3): 3,435 patients were included (Table 1). - Patients who were younger, with no prior therapy, or ECOG 1. breast cancer cases and has poor Variable All ( N=777) - bevacizumab combined with paclitaxel in 1L post-metastasis were more likely to receive taxane/anthracycline in 1L (Table 3).

2 Since patients do not benefit from hormonal or trastuzumab-based Hospital type, N (%) - capecitabine in 2L - Patients who had ECOG 1, no comorbidities, failed therapies, cytotoxic chemotherapy remains the main Treatment Table 1: Patient Selection and Demographic Characteristics University/Teaching Hospital 381 (49%) - eribulin in 3L taxane/anthracycline Treatment in 1L, or received monotherapy modality for metastatic TNBC (mTNBC).1 Risk of relapse remains Variable General 187 (24%) in 1L were more likely to receive capecitabine in 2L (Table 3). high even after chemotherapy. Comprehensive Cancer Center 160 (21%) Patients with ECOG 2 commonly received capecitabine as Patient Selection Other 49 (6%) monotherapy across all lines post-metastasis (Figure 3).

3 Common recommended treatments include taxanes and/or All patients with TNBC, N 4,844 Doctor specialty, N (%). anthracyclines in first-line (1L) and agents such as capecitabine or Female patients age 18 years, N 4824 Medical Oncology 633 (81%) Table 3: Logistic Regression Analyses to Determine gemcitabine in progressed or relapsed Patients with current stage IV diagnosis, N 3,440 Figure 3: Frequently Received Treatments in Each Line by ECOG. Hematology/Oncology 100 (13%) Predictors of Receipt of Commonly Used Chemotherapy Although Functional Status is a critical Treatment selection factor Patients with known stage at diagnosis, N 3,435 Gynecological Oncology 42 (5%). Scores when seeking to balance chemotherapy benefits and costs, there is Other 2 ( ) 1st line of therapy Variables Category Odds Ratio p-value Demographic Characteristics All (N= 3,435) (95% CI).

4 Lack of information on real-world Treatment Patterns by Functional Country, N (%). Years since specialized in cancer Status . Mean/Median (SD) ( ) Model 1: Predictors of Receipt of Taxanes/Anthracyclines in 1L. France 756 (22%) Age ( ) < Bevacizumab/ Paclitaxel Spain 756 (22%) ECOG ECOG>1 vs. ECOG<=1 ( ) < Italy 727 (21%). OBJECTIVE Germany 632 (18%) ECOG Scores by Treatment Line Prior combination Yes vs. No ( ) < Paclitaxel therapy UK 564 (17%) 60% of patients had ECOG score=1 in 1L post-metastasis (Figure To describe European mTNBC Treatment Patterns by Eastern Age Model 2: Predictors of Receipt of Capecitabine in 2L. Cooperative Oncology Group (ECOG) scores. 1). Mean/Median (SD) ( ) Capecitabine ECOG ECOG>1 vs. ECOG<=1 ( ) < Gender, N (%) Additional treatments lines were Associated with worsening ECOG Prior combination Yes vs.

5 No ( ) < Female 3,435 (100%) scores (Figure 1). therapy METHODS. Docetaxel Race, N (%) Comorbidities Yes vs. No ( ) Caucasian 3,122 (92%) Receipt of Yes vs. No ( ) < Asian 102 (3%) Figure 1. ECOG Scores by Current metastatic Treatment Line anthracycline/. The analysis used the Ipsos Healthcare's Global Oncology Monitor African American 98 (3%). Eribulin Database, a large cross-sectional patient record database extracted taxane in prior line Other/unknown 69 (2%). by physicians (2014-2016). Body Mass Index (BMI) 100%. 2% 1% 1% ECOG 0-1 (N= 1609) ECOG 2 (N= 336). The database was used to assess routine Treatment Patterns of Mean/Median (SD) ( ) 90% 16%. LIMITATIONS. 19% 2nd line of therapy mTNBC in EU5 (France, Germany, Italy, Spain and United Kingdom Concomitant conditions, N (%) 80% 34%.)

6 Hypertension 982 (29%). Percentage of Patients [UK]). 70%. Cross-sectional design due to nature of data ( , longitudinal Diabetes 680 (20%) 3 Adult female TNBC patients 18 years with stage IV disease Depression/Anxiety 366 (11%). 60%. 2. Capecitabine analyses could not be performed). 60%. ( metastatic disease) at the time of data extraction were identified Cardiovascular disease 301 (9%) 50% 63% 1 Limited to EU5 population and may not be generalizable to other (Table 1). Patients who had unknown disease stage were excluded. Thyroid disorders 277 (8%) 40% 55% 0 Eribulin countries. Pulmonary disorders 225 (6%) 30%. Regimens across 1L, second-line (2L) and third-line (3L) were Obesity 173 (5%) 20%. identified based on their current regimen at time of data Smoking Status , N (%).

7 Paclitaxel extraction. 10% 23%. 16%. 10%. CONCLUSIONS. No 2,167 (63%) 0%. Statistical Analyses: Yes 374 (11%) Mean ECOG (SD) 1st line 2nd line 3rd line Docetaxel We described EU5 real-world Treatment Patterns in mTNBC. ( ) ( ) ( ). Quit 821 (24%). - Frequency and percentage were used to report categorical Unknown 73 (2%). Treatment lines Patients most often receive taxane-containing Treatment as variables and mean and standard deviation were used for metastatic sites, N (%). Vinorelbine Alternative agents ( , capecitabine and eribulin). continuous variables. Lung 1,922 (56%) were commonly given as successive lines are used. Treatment Selection by ECOG Scores - Predictors of receipt of anthracycline/taxane in 1L and Liver 1,796 (52%) ECOG 0-1 (N= 886) ECOG 2 (N= 229).

8 An association between ECOG scores and Treatment selection capecitabine in 2L were identified via logistic regression. Bone 1,474 (43%) Fewer patients received combination therapy with each successive Lymph nodes 1,320 (38%) line (Figure 2). 3rd line of therapy post-metastasis was identified: Other sites 354 (10%). Patient with ECOG 1 more frequently received combination - Patients with better ECOG were treated more aggressively with Stage at diagnosis, N (%). therapy in 1L post-metastasis vs. patients with ECOG>1 (Figure 2). Eribulin combination therapy versus patients with worse ECOG (Figure Stage I 109 (3%) 2). Stage II 498 (14%). Stage III 509 (15%) Capecitabine - Capecitabine monotherapy was more common among patients Stage IV 2,319 (67%) Figure 2.

9 Patients with Combination Therapy by Line and with worse ECOG versus those with better ECOG (Figure 3). ECOG Scores, N (%) ECOG Score - Patients who received more aggressive combination Treatment Vinorelbine 0 659 (19%) Patients with Combination Therapy by Line and ECOG Score in 1L, received taxane/anthracycline in 1L or 2L or received 1 2,072 (60%). 2 644 (19%). 60% capecitabine in 2L after failing prior taxane/anthracycline SPONSORSHIP. Gemcitabine 3 48 (1%). Treatment had better ECOG (Table 3). 50%. This research was funded by Celldex Therapeutics Inc, Hampton, NJ Mean/Median (SD) ( ) Carboplatin/ Gemcitabine Percentage of Patients 40%. ECOG 0-1 (N=223) ECOG 2 (N=123). REFERENCES. Physician Characteristics (Table 2) 30% 1. Anders CK, Carey LA.

10 Biology, metastatic Patterns , and Treatment of The 3,435 patients were treated by 777 physicians (Table 2). 20%. Patients receiving taxanes/anthracyclines in 1L or 2L had better patients with triple-negative breast cancer. Clin Breast Cancer. 2009;. - 49% of physicians were Associated with a university or teaching ECOG scores vs those not receiving taxanes/anthracyclines ( vs 9(suppl 2): S73-S81. hospital. 10% , p< ). 2. Andre F, Zielinski CC. Optimal strategies for the Treatment of metastatic Among patients who failed prior taxane/anthracycline Treatment , triple-negative breast cancer with currently approved agents. Annals of - The most common specialty was medical oncology (81%) 0% Oncology. 2012; 23(suppl 6):vi46-vi51. those receiving capecitabine in 2L had better ECOG scores - Physicians have been specializing in oncology for a median of ECOG 0-1 ECOG 2 ECOG 0-1 ECOG 2 ECOG 0-1 ECOG 2.