Search

The clinical utility of tumor-infiltrating lymphocytes (TIL) is actively being investigated in breast cancer (BC). It is unclear whether TIL spatial location and organization in tertiary lymphoid structures (TLS) have an impact on prognosis. Additionally, the significance of PD-1 and PD-L1 expression is being debated due to conflicting data from several studies. We hypothesize that the presence, extent and spatial location of multiple immune biomarkers, reflecting ongoing immune responses, will be consistently associated with a good prognosis in highly infiltrated BC [triple-negative (TNBC) and HER2+]. The relationship between these immune biomarkers and clinical outcome was examined in the TNBC and HER2+ cohorts of node-positive BC patients enrolled in the BIG 02-98 adjuvant phase III trial with available material for immunohistochemical (IHC) labeling (N=113 and N=136, respectively). HER2+ patients did not receive trastuzumab. Dual IHC staining was performed on full-face consecutive tissue sections. Scoring was independently performed by two pathologists, blinded to the clinical data, and included: global, intratumoral and stromal TIL and TLS, assessed on CD3/CD20 slides; the percentage and location of PD-1 and PD-L1 expression, assessed on PD-1/PD-L1 slides. TIL were considered as a categorical variable with different cut-offs used for each parameter and for each cohort (TNBC and HER2+). Invasive disease-free survival (I-DFS) and overall survival (OS) were analyzed (median follow-up: 10 years). Cox proportional hazard models were used for survival analyses. The TNBC cohort revealed an association between global TIL and outcome [adjusted hazard ratio (HR) for I-DFS: 0.27 (0.15-0.51); OS: 0.26 (0.13-0.53)]. Similar results were observed for stromal and intratumoral TIL. PD-L1 expression within TLS was an independent predictor of OS, after adjustment for tumor size and age [HR: 0.30 (0.09-0.99)]. Multivariate analysis reveals this effect was principally driven by high stromal TIL (>17.5% based on CD3/CD20 assessment) (χ2 OS: p=0.009). In contrast, no significant prognostic associations were found in the overall HER2+ cohort. However high T cell TIL were associated with improved I-DFS and OS in the ER-/HER2+ group [I-DFS: 0.34 (0.14-0.80); OS: 0.32 (0.12-0.86)] and stromal TIL were associated with improved I-DFS in the ER+/HER2+ group [HR: 0.29 (0.09-0.94)] (univariate analyses). No significant associations between the number of TLS nor the expression of PD-1 with outcomes were observed in either cohorts. The presence of PD-L1+ TLS, driven by high baseline TIL, was associated with an excellent prognosis in node-positive TNBC. This observation might reflect specific immune activities taking place in these mini lymph node-like structures adjacent to the tumor bed where specific antitumor memory immune responses could be generated. No different prognostic impact was observed when analyzing TIL spatial location. Although the statistical power of the study might be limited, in line with previous findings our data reveal that, among the immune parameters evaluated, TIL are the strongest predictor of outcome in TNBC, while PD-L1+ TLS could be a new and important parameter that requires further investigation. Citation Format: Solinas C, de Wind A, Van den Eynden G, Ameye L, Garaud S, De Silva P, Boisson A, Noel G, Langouo Fontsa M, Buisseret L, de Azambuja E, Francis PA, Di Leo A, Crown JP, Sotiriou C, Larsimont D, Paesmans M, Piccart-Gebhart M, Willard-Gallo K. Immune parameters associated with survival in triple negative and HER2-positive breast cancer patients with 10 years of follow-up [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr PD5-09.

Background: Optimal endocrine therapy for premenopausal pts with early HR+ BC may depend on complete estrogen suppression with GnRH analog, which is crucial when using concurrent aromatase inhibitors (AIs). SOFT-EST is a prospective substudy of the phase 3 SOFT trial aiming to describe estradiol (E2), estrone (E1) and estrone sulphate (E1S) during the first 4 years (y) of monthly Trip+E/T and to assess if there were suboptimally estrogen suppressed (SES) pts in the E+Trip group. Secondary objectives included associations of baseline (BL) factors with SES, early SES with later SES, and SES with disease-free survival (DFS; exploratory objective). Methods: Patients from select centers who consented and enrolled in SOFT, selected Trip as ovarian function suppression method, and were randomized to E+Trip or T+Trip were eligible for SOFT-EST until the accrual goal (120 pts: 90 E+Trip; 30 T+Trip). Prem status for SOFT eligibility was based on local E2. Blood sampling timepoints were 0, 3, 6, 12, 18, 24, 36 & 48 months (m) until Trip stopped. Serum estrogens were measured centrally by high specificity/sensitivity GC/MSMS and were not available during the study. For 4y analyses, SES was defined as E2 levels >2.72 pg/mL in ≥2 post-BL samples (E2 levels not consistent with postmenopausal (PM) status on AIs [Smith IE, JCO 2006]), or vaginal bleeding >3m after Trip start, or pregnancy. We explored 2 additional cutoffs: >10 pg/mL (clearly inconsistent with PM status on AIs) and >20 pg/mL (inconsistent with GnRH analog-related PM status). The analysis is intention-to-treat based on E/T assignment; as-treated analyses are forthcoming. Results: From Mar 2009 to Jan 2011,109 pts (E/T=83/26) started Trip and had ≥2 samples drawn. In pts assigned E+Trip, median reductions from BL in E1, E2 and E1S were >95% at all timepoints and significantly lower than in T+Trip. Post-BL E2 geometric mean ranged 0.8-1.3 pg/mL in E+Trip and 16.5-18.3 pg/mL in T+Trip. 21 (25%), 11 (13%) and 6 (7%) pts assigned to E+Trip had E2>2.72, >10, and >20 pg/mL in ≥2 post BL samples or vaginal bleeding (n=3), respectively. Early SES [(≥1 E2 value >2.72 pg/mL or vaginal bleeding in the firsty] predicted later SES [≥1 E2 value >2.72 or vaginal bleeding thereafter (n=1); p Conclusions: Most pts on E+Trip had a profound E2 drop consistent with postmenopausal status on AI, but >20% assigned to E+Trip had ≥2 E2 values >2.72 pg/mL and 4% had vaginal bleeding, with those having higher E2, lower FSH/LH at BL being at higher risk. SES at 12m predicted subsequent SES. Few DFS events limit the ability to assess clinical relevance of SES with disease outcomes. BL characteristicsN-109Prior chemo60 (55%)Amenorrhea39 (36%)Age Citation Format: Bellet M, Gray K, Francis P, Láng I, Ciruelos E, Lluch A, Ángel Climent M, Catalán G, Avella A, Bohn U, González-Martin A, Zaman K, Ferrer R, Azaro A, Rajasekaran A, De la Peña L, Fleming G, Regan MM. Estrogen levels in premenopausal patients (pts) with hormone-receptor positive (HR+) early breast cancer (BC) receiving adjuvant triptorelin (Trip) plus exemestane (E) or tamoxifen (T) in the SOFT trial: SOFT-EST substudy final analysis [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P4-14-01.

Background Radiation therapy (RT) after breast conserving surgery (BCS) is the current standard of care for patients with early stage breast cancer. However, individual absolute recurrence risks and hence benefits of RT vary substantially. A study showed significant association between local recurrence (LR) risk and PAM50-defined intrinsic subtypes and Risk of Recurrence scores (ROR).1 The objective of EXPERT, a co-lead study of Breast Cancer Trials-Australia & New Zealand (BCT-ANZ), and Breast International Group (BIG), is to optimize local therapy for early breast cancer through precise individualized quantification of LR risk to identify patients for whom RT after BCS may be safely omitted. Trial design This is a randomized, non-inferiority, phase III study of women who plan to receive adjuvant endocrine therapy for Prosigna (PAM50)-defined luminal A breast cancer with ROR ≤60 resected by BCS. Women are randomized to receive adjuvant whole breast RT and endocrine therapy or endocrine therapy alone and followed-up for 10 years after randomization. Major eligibility criteria Females aged ≥50 years; histologically confirmed invasive breast carcinoma ≤2 cm, grade 1 or 2, ER and PgR ≥10%, HER2-negative and node-negative; treated by BCS with negative margins for invasive carcinoma and associated DCIS; Prosigna (PAM50)-defined Luminal A subtype and ROR ≤60; and plan to receive adjuvant endocrine therapy. Specific aims Primary: To determine if omission of RT is not inferior to RT in terms of LR-free interval after BCS. Secondary: To evaluate the impact of omission of RT on regional, local-regional and distant recurrence-free interval; disease-free survival (DFS); invasive DFS; overall survival; salvage RT or mastectomy rate; toxicity; endocrine therapy adherence; patient reported outcomes; and health economic outcomes. Statistical methods An estimated 5-year LR rate in the target population is expected to be 1% with RT. A rate of 4% is considered non-inferior as a worthwhile trade-off against RT toxicity. Using O'Brien-Fleming boundary for rejecting non-inferiority, 29 LR events are required for final analysis expected 8 years after the first patient is randomized. Two interim analyses will be conducted after 10 and 21 events. If the stratified log-rank test statistic exceeds the upper boundary at interim or final analysis, the hypothesis of non-inferiority will be rejected and it will be concluded that no RT is inferior to RT. Accrual: Target (1170), actual: 82 (June 2018) The study was activated in Australia in August 2017, with global activation planned for Q4 2018. Recruitment is expected to be completed in 4.5 years. Contact information Professor Boon Chua, UNSW Sydney and Prince of Wales Hospital, NSW, Australia; email boon.chua@health.nsw.gov.au; T +61 2 49255239. Registration: NCT02889874 References Fitzal F, Filipits M, Fesl C, et al. Predicting local recurrence using PAM50 in postmenopausal endocrine responsive breast cancer patients. JCO 2014;32(15 suppl):1008. Citation Format: Chua BH, Gray K, Krishnasamy M, Regan M, Zdenkowski N, Loi S, Mann B, Forbes JF, Wilcken N, Spillane A, Martin A, Badger H, Jafari S, Fong A, Mavin C, Corachan S, Arahmani A, Martinez J-L, Francis P. Examining personalized radiation therapy (EXPERT): A randomised phase III trial of adjuvant radiotherapy vs observation in patients with molecularly characterized luminal A breast cancer [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr OT2-04-03.

Background: In premenopausal women with breast cancer any treatment that causes abrupt, premature ovarian failure increases the risk of sexual problems. Randomized-controlled trials in this population reported a worsening in sexual functioning over time irrespective of adjuvant endocrine treatment. We investigated key symptoms related to endocrine therapy, depression and age as predictors of sexual problems in premenopausal women with early breast cancer treated in the IBCSG TEXT/SOFT trials over the first two years of endocrine therapy. Methods: A subset of patients (pts) enrolled by centers with English as primary language to TEXT (1027 of 2672 pts) and SOFT (1260 of 3066 pts) completed a questionnaire consisting of global and symptom-specific quality of life indicators, the CES-Depression (CES-D) and the MOS- Sexual Problems (MOS-SP) measures at baseline, 6, 12 and 24 months. The analysis considered 5 cohorts of pts according to chemotherapy use (yes/no), trial (SOFT/TEXT) and endocrine treatment assignment (tamoxifen alone [T], T or exemestane [E] with ovarian function suppression [OFS]). Mixed modeling was used to test the effect of the following on changes in sexual problems (MOS-SP total score) over two years: changes in treatment-induced symptoms (hot flushes, vaginal dryness, sleep disturbances, bone/joint pain, troubled by weight gain, tiredness, nausea/vomiting) from baseline to 6 months; depression at 6 months; and age at randomization. The model included severity groups of symptoms, depression (all dichotomized by median) and age (< 40 vs ≥40 years), 5 cohorts, time points (6, 12, 24 months), baseline covariates, and interactions of symptoms, timepoints and cohorts. Results: Overall across cohorts, pts with more severe worsening of vaginal dryness and sleep disturbances at 6 months reported a greater increase in sexual problems at all timepoints (p Conclusion: Among several key symptoms related to endocrine therapy, only vaginal dryness and sleep disturbances significantly predicted sexual problems during the first two years in pts who received adjuvant endocrine therapy with or without chemotherapy. Depression predicted sexual problems only in the cohort of pts who received combined endocrine treatment without chemotherapy. Early identification of vaginal dryness, sleep disturbances and depression is important for timely and tailored interventions. Citation Format: Ribi K, Luo W, Burstein HJ, Naughton MJ, Chirgwin J, Ansari RH, Walley BA, Salim M, van der Westhuizen A, Abdi E, Francis PA, Budman DR, Kennecke H, Harvey VJ, Giobbie-Hurder A, Fleming GF, Pagani O, Regan MM, Bernhard J. The effects of treatment-induced symptoms, depression and age on sexuality in premenopausal women with early breast cancer receiving adjuvant endocrine therapy [abstract]. In: Proceedings of the 2016 San Antonio Breast Cancer Symposium; 2016 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2017;77(4 Suppl):Abstract nr P2-09-09.

Background Pre-clinical data and initial results from clinical studies point to the added benefit of CDK4/6 inhibition when combined with anti-HER2 tx. The current study is designed to evaluate the added benefit of palbociclib when given in combination with anti-HER2 and endocrine tx maintenance in the 1st†line setting of metastatic HER2+HR+ breast cancer. Trial design PATINA is an international, open-label, pivotal Phase III study. Primary objective is to demonstrate that the combination of palbociclib with anti-HER2 plus endocrine tx is superior to anti-HER2 plus endocrine tx in prolonging PFS. Sample size is 496 pts. The study starts after completion of 6-8 cycles of chemotherapy-containing anti-HER2 tx for metastatic breast cancer in the 1st line setting. Pts are eligible provided they are without evidence of disease progression by local assessment (i.e. CR, PR or SD). To account for the need for less intense tx regimens for a subset of pts diagnosed with HER2+ER+ disease, clinicians may recommend the combination of trastuzumab with either a taxane or vinorelbine prior to study initiation. Clinicians might also choose a non-pertuzumab option for pts previously treated with pertuzumab in the neo(adjuvant) setting. Secondary objectives include measures of tumor control (OR, CBR, DOR), OS, safety and QOL. The translational science main objective is to compare PFS estimates according to PIK3CA mutation status assessed by cfDNA analysis. Endocrine tx options are AI or fulvestrant. Premenopausal pts must receive ovarian suppression. The study has a 90% power to detect a hazard ratio of 0.667 in favor of the palbociclib arm. Pts approached to participate in AFT-38 will be asked to indicate on the informed consent forms whether remaining biospecimens and clinical data from the control arm of the study can be shared with the Mastering Breast Cancer (MBC) Initiative. The overarching purpose of the MBC is to create a mechanism for understanding the natural history of metastatic breast cancer by cataloguing longitudinally studied tumor-specific markers and treatment effects. ClinicalTrials.gov Identifier: NCT02947685 Citation Format: Metzger-Filho O, Mandrekar S, Loibl S, Ciruelos E, Gianni L, Lim E, Miller K, Huang C, Koehler M, Francis P, Valagussa P, Goel S, Prat A, Goetz M, Loi S, Krop I, Carey L, Lanzillotti J, Winer E, Tripathy D, DeMichele A. PATINA: A randomized open label phase III trial to evaluate the efficacy and safety of palbociclib + anti HER2 therapy + endocrine therapy vs anti HER2 therapy + endocrine therapy after induction treatment for hormone receptor positive, HER2 positive metastatic breast cancer [abstract]. In: Proceedings of the 2017 San Antonio Breast Cancer Symposium; 2017 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2018;78(4 Suppl):Abstract nr OT3-05-07.

Background It is uncertain if the addition of OFS to adjuvant endocrine therapy with T improves outcomes in premenopausal HR+ BC. The SOFT trial was designed to determine the value of OFS in women who remain premenopausal and are treated with T (OFS question) and also to test whether an AI improves outcome in premenopausal women with HR+ BC treated with OFS (AI question). In the combined analysis of the TEXT and SOFT trials (AI question), the group treated with the AI exemestane+OFS had a significantly better disease-free survival (DFS) than those with T+OFS. This abstract presents the results of the OFS question (n=2,045). Patients and Methods Between November 2003 and January 2011, the SOFT phase 3 trial randomized 3066 premenopausal women with HR+ BC to 5 years of adjuvant endocrine therapy with exemestane+OFS versus T+OFS versus T alone, with OFS initiated by choice of GnRH agonist triptorelin, oophorectomy or ovarian irradiation. Prior chemotherapy was allowed, provided women had confirmed premenopausal estradiol levels within 8 months of completing chemotherapy. The primary end point was DFS which included invasive local, regional, distant and contralateral breast events, second non-breast malignancies and deaths without prior cancer. Because of a lower-than expected overall event rate and to ensure results within a reasonable time-frame, a protocol amendment in 2011 changed the originally event-driven analysis plan. The amendment (1) designated the comparison of T+OFS versus T alone as the primary analysis for SOFT (n=2045), recognizing that the statistical power for the original three pairwise comparisons would be substantially reduced and (2) specified that the primary analysis of T+OFS versus T be undertaken when median follow-up was at least 5 years. The comparison would be tested at the 2-sided 0.05 level with no interim analysis. Based on the projected number of events at the lower rates, the estimated power to detect hazard ratios of 0.75, 0.70, and 0.66 for the comparison would be 52%, 69%, and 80%. Results The SOFT trial completed planned patient enrollment with an international collaboration involving 426 centres from BIG and NABCG led by the International Breast Cancer Study Group (IBCSG). Numbers of patients randomized, randomization strata, and pt age are summarized in the Table. The database lock for the primary analysis of the OFS question will occur in September 2014 and the final analysis will be completed by October 15, 2014. SOFT Patients (%)Tamoxifen alone1021 (33%)Tamoxifen+OFS1024 (33%)Exemestane+OFS1021 (33%)* Prior Chemotherapy54%Lymph node positive35%Median age (% < 40 years)43 years (30%)*Patients randomized to Exemestane+OFS are not part of the current analysis Conclusions We will present the primary analysis of outcomes and toxicities by treatment for women randomized to receive T+OFS versus T and address the value of OFS in addition to T as adjuvant endocrine therapy for premenopausal women with HR+ BC. Citation Format: Aron Goldhirsch, Richard D Gelber, Prudence A Francis, Meredith M Regan, Gini F Fleming, Istvan Lang, Eva M Ciruelos, Meritxell Bellet, Herve Bonnefoi, Miguel A Climent, Lorenzo Pavesi, Harold J Burstein, Silvana Martino, Nancy E Davidson, Charles E Geyer Jr, Barbara A Walley, Robert E Coleman, Pierre Kerbrat, Manuela Rabaglio-Poretti, Alan S Coates. Randomized comparison of adjuvant tamoxifen (T) plus ovarian function suppression (OFS) versus tamoxifen in premenopausal women with hormone receptor-positive (HR+) early breast cancer (BC): Analysis of the SOFT trial [abstract]. In: Proceedings of the Thirty-Seventh Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2014 Dec 9-13; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2015;75(9 Suppl):Abstract nr S3-08.

Adjuvant therapy decisions for premenopausal ER+ve breast cancer are challenging. An analysis from TEXT found that therapy varied by geography.1 Approximately half the women with node negative tumors enrolled from North America received optional adjuvant chemotherapy, in addition to protocol ovarian suppression plus oral endocrine therapy, while only one quarter of women with node negative tumors enrolled from Italy received chemotherapy. For women with node positive tumors, the use of chemotherapy was almost universal (93%) in North America, while 27% enrolled from Italy did not receive chemotherapy. Use of multigene assays may reduce future use of chemotherapy in ER+ve breast cancer, however limited information exists for outcomes of young women with low risk assay results, treated with tamoxifen alone. Among 6,693 women enrolled in MINDACT, only 44 women were under age 35 with low genomic risk.2 Only 58 women aged 40 or under were included in the Tailor-X Recurrence Score 0-10 cohort and not all received tamoxifen alone.3 Chemotherapy may provide an indirect endocrine effect in premenopausal ER+ve breast cancer, due to suppression of ovarian estrogen production. The EBCTCG meta-analysis published in 1998 reported that 5 years of adjuvant tamoxifen was effective, regardless of age, menopausal status or chemotherapy.4 Adjuvant ovarian ablation is effective for women under 50 as a sole adjuvant therapy,5 but the value of ovarian suppression/ablation in premenopausal women with ER+ve tumors who also receive tamoxifen, with or without adjuvant chemotherapy, has been uncertain.6 SOFT randomized premenopausal women with hormone receptor-positive breast cancer to receive 5 years of tamoxifen or tamoxifen plus ovarian suppression or exemestane plus ovarian suppression.7 The initial results, after median 5.6 years follow-up, did not show a significant benefit from addition of ovarian suppression to tamoxifen in the overall premenopausal population. SOFT was stratified by use of prior chemotherapy and in the cohort who did not receive chemotherapy, women allocated tamoxifen alone had few recurrences and virtually no distant recurrences during the first 5 years. However, for patients in SOFT deemed at sufficient risk to receive chemotherapy and who remained premenopausal, the addition of ovarian suppression to tamoxifen did reduce risk of invasive recurrence during the first 5 years, and the use of exemestane with ovarian suppression further reduced recurrence. The joint analysis of TEXT and SOFT found that exemestane plus ovarian suppression significantly improved DFS compared with tamoxifen plus ovarian suppression.8 The absolute benefit of endocrine therapies that included ovarian suppression in SOFT and TEXT, as compared with tamoxifen alone, was most striking in women under age 35, although women with a high risk of recurrence from a combination of clinical-pathologic factors (composite risk) could also derive benefit.9 References: 1. Regan et al. Ann Oncol 2008; 19: 1231-1241 2. Cardoso et al. NEJM 2016; 375:718-29 3. Sparano et al. NEJM 2015; 373:2005-14 4. EBCTCG. Lancet 1998; 351:1451-67 5. EBCTCG. Lancet 1992; 339:1-15 6. EBCTCG. Lancet 2007; 369:1711-23 7. Francis et al. NEJM 2015; 372: 436-446 8. Pagani et al, NEJM 2014; 371:107-18 9. Regan et al. J Clin Oncol 2016; 34: 2221-31 Citation Format: Francis PA. Adjuvant endocrine therapy for premenopausal ER+ breast cancer [abstract]. In: Proceedings of the 2017 San Antonio Breast Cancer Symposium; 2017 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2018;78(4 Suppl):Abstract nr CS1-1.

Background: The combined results of TEXT and SOFT, after 5.7 years median follow-up, found adjuvant E+OFS significantly improved disease-free survival (DFS) vs T+OFS in premenopausal women with HR+ BC (Pagani et al, NEJM 2014). Follow-up was immature for overall survival (OS). We report a planned update with visit cut-off of 31Dec16 after 9 years median follow-up. Methods: TEXT and SOFT enrolled premenopausal women with HR+ early BC from Nov 2003 to Apr 2011 (2660 TEXT, 3047 SOFT in the intention-to-treat populations). TEXT randomized women within 12wk of surgery to 5 yrs E+OFS vs T+OFS; chemotherapy (CT) was optional and concurrent with OFS. SOFT randomized women to 5 yrs E+OFS vs T+OFS vs T alone, within 12wk of surgery if no CT planned, or within 8mo of completing (neo)adjuvant CT after premenopausal status was (re-)established. OFS was by choice of 5yr GnRH agonist triptorelin, oophorectomy or ovarian irradiation. Both trials were stratified by CT use. The primary endpoint was DFS: randomization until invasive local, regional, distant recurrence or contralateral breast; invasive second malignancy; death. Secondary endpoints included invasive breast cancer-free interval (BCFI), distant recurrence-free interval (DRFI) and OS. Stratified Cox models estimated hazard ratios; Kaplan-Meier method estimated 8yr endpoint rates. NCT00066703/NCT00066690. Results: DFS for patients assigned E+OFS (n=2346) continued to be significantly improved over T+OFS (n=2344): 8yr DFS was 86.8% vs. 82.8%. The 8yr BCFI was improved by 4.1% (89.3% vs 85.2%) and 8yr DRFI by 2.1% (91.8% vs 89.7%). There was no difference in OS in patients assigned E+OFS vs T+OFS: 93.4% vs 93.3% OS at 8yrs. For 1996 women without CT there have been 45 deaths, with 98% OS at 8yrs with both treatments. EndpointN. EventsHazard Ratio (95% CI) E+OFS vs T+OFSDFS7200.77 (0.67-0.90); P Overall toxicity was not significantly worse with E+OFS than with T+OFS (32% vs 31% grade 3-4 targeted AEs). Hot flashes, musculoskeletal symptoms and hypertension were the most frequent targeted grade 3-4 AEs. Overall, 15% of patients stopped all protocol-assigned treatment early. Patients assigned E+OFS had increased risk of assigned oral endocrine therapy cessation (25% vs 19% for patients assigned T+OFS by 4yrs) but not of triptorelin cessation (18% vs 19% by 4yrs, respectively). Conclusions: After 9 yrs median follow-up, adjuvant E+OFS, as compared with T+OFS, shows a sustained reduction of the risk of recurrence but did not improve overall survival. As in postmenopausal women, oncologists need to consider potential absolute benefits and properly select patients at sufficient risk for recurrence for whom E+OFS seems indicated. Follow-up continues, which will further clarify the effect of E+OFS for safety, late recurrence and overall survival. Citation Format: Pagani O, Regan MM, Fleming GF, Walley BA, Colleoni M, Láng I, Gomez HL, Tondini C, Burstein HJ, Goetz MP, Ciruelos EM, Stearns V, Debled M, Martino S, Geyer Jr CE, Pinotti G, Coates AS, Goldhirsch A, Gelber RD, Francis PA. Randomized comparison of adjuvant aromatase inhibitor exemestane (E) plus ovarian function suppression (OFS) vs tamoxifen (T) plus OFS in premenopausal women with hormone receptor positive (HR+) early breast cancer (BC): Update of the combined TEXT and SOFT trials [abstract]. In: Proceedings of the 2017 San Antonio Breast Cancer Symposium; 2017 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2018;78(4 Suppl):Abstract nr GS4-02.

Background First results of the TEXT and SOFT international phase III trials were practice-changing, indicating that: i) 5y adjuvant exemestane+ovarian function suppression (E+OFS) reduces recurrence risk relative to tamoxifen(T)+OFS or to T alone, ii) T+OFS reduces recurrence risk vs T in women who are at sufficient risk to warrant chemotherapy (CT) and remain premenopausal afterwards, and iii) T alone remains appropriate for some premenopausal women. However, median follow-up (FU) was only 5.5y and Trial Design and Aims Premenopausal women had invasive early breast cancer (BC) assessed as ≥10% ER and/or PgR. SOFT was designed to determine the value of adding OFS to T, and the role of E+OFS in two cohorts: women who remained premenopausal after completion of neo/adjuvant CT, and women for whom adjuvant T alone was considered suitable treatment. SOFT compares 5y of T to T+OFS or E+OFS. OFS was GnRH analog triptorelin x5y, oophorectomy or ovarian irradiation. Median age was 43y; 35% had N+ disease. 53% enrolled after prior neo/adjuvant CT. TEXT was designed to determine the role of adjuvant E in premenopausal women receiving OFS from the start of adjuvant therapy, comparing 5y of E+OFS vs T+OFS. Patients enrolled at start of all adjuvant therapy; 60% had CT concurrent with triptorelin after entry. Median age was 43y; 48% had N+ disease. Secondary objectives include effects on OS, late side effects of early menopause and late toxicities. Accruals TEXT: 2672 women, Nov03-Mar11 SOFT: 3066 women, Dec03-Jan11 Statistical Methods The primary endpoint, invasive disease-free survival, is time from randomization to invasive local, regional, or distant relapse, contralateral BC, second non-BC malignancy, or death. Secondary endpoints are BC-free interval, distant recurrence-free interval and OS. Primary results were reported in 2014, after ∼5.5y median FU; 30% pts were still on 5y treatment and >90% continued in FU. Long-term FU Updated results are planned for FU through Dec16, with ∼8y median FU. Pts finished 5y treatment by Apr16. Yearly visits continue; data collection includes weight, performance status, menstrual status, pregnancy attempts, GYN procedures, late AEs (cardiovascular, bone fracture), extended adjuvant therapy, invasive recurrence at first and subsequent sites, second non-BC malignancy, in situ cancers, OS. FU through 2020 is planned, for min and median FU of 10 and 12y, roughly doubling the numbers of endpoints events since the first report. This will be critical to determine whether short-term treatment benefits persist for late recurrence, improve power to detect treatment effects on distant recurrence and OS endpoints with lower event rates occurring later in FU, and define associated late toxicities and side effects of early menopause. A consortium to fund long-term FU is being pursued. Citation Format: Francis PA, Fleming GF, Regan MM, Pagani O, Walley BA, Price KN, Coates AS, Goldhirsch A, Gelber R. Long-term follow-up of TEXT and SOFT trials of adjuvant endocrine therapies for premenopausal women with HR+ early breast cancer [abstract]. In: Proceedings of the 2016 San Antonio Breast Cancer Symposium; 2016 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2017;77(4 Suppl):Abstract nr OT3-02-03.

Background: The SOFT and TEXT randomized phase 3 trials address two primary questions for endocrine treatment of premenopausal women with hormone receptor-positive breast cancer. 1) In combination with ovarian function suppression (OFS), does an aromatase inhibitor (exemestane, E) improve outcome compared with tamoxifen (T)? 2) Does addition of OFS to T improve outcome compared with T alone? Trial Designs: SOFT compares 5y of T to OFS+T or OFS+E. OFS can be GnRH analog (triptorelin) × 5y, oophorectomy or ovarian irradiation. Median age was 43y (11% TEXT compares 5y of OFS+T to OFS+E. Patients were enrolled prior to CT (if planned). Median age was 43y (9% Major Eligibility Criteria – Premenopausal, confirmed by estradiol levels– ER≥10% and/or PgR≥10%– Invasive early breast cancer Specific Aims: Evaluate the role of aromatase inhibitors and the addition of OFS to T in this population. Statistical Methods (amended 2011): The primary analysis will be intention-to-treat of all randomized patients. The primary endpoint, invasive disease-free survival (DFS), is defined as time from randomization to invasive local, regional, or distant relapse, contralateral breast cancer, appearance of a second non-breast malignancy, or death. DFS will be compared using a 2-sided stratified logrank test with an overall experiment-wise alpha level equal to at most 0.05. In the original protocol, anticipated 5y DFS was 67% with T alone, 74% with OFS+T, and 79.8% with OFS+E, and required a total of 396 DFS events in TEXT and 783 DFS events in SOFT to reach 80% power. Because the enrolled population had more favorable characteristics and the event rates were lower than anticipated (2%/yr vs 8%/yr), the revised analysis plan is ‘time-driven’ rather than ‘event-driven.’ The comparison of OFS+E to OFS+T across both trials (n = 4717) is planned at median follow-up (MFU) >5y. The estimated power to detect a 20%, 25%, or 30% reduction in the hazard with OFS+E vs OFS+T is 63%, 84%, and 95%, respectively. The comparison of OFS+T to T alone is planned at 5y MFU (SOFT, n=2045). The estimated power to detect a 20%, 25%, 30%, or 33.5% reduction in the hazard is 34%, 52%, 69%, and 80%, respectively. Accruals SOFT Target: 3000; Final: 3066 TEXT Target: 2639; Final: 2672 Enrollment 2003–2011; primary analyses expected late 2013/early 2014. Related Research Quality of Life (QL) component evaluates QL, menopausal symptoms and sexual impairment. TEXT Translational Research investigates patient and tumor features that may contribute to treatment effectiveness and side effects. TEXT-Bone investigates changes in bone mineral density and the role of serial serum markers of bone remodeling as predictors of bone side effects. Co-SOFT evaluates changes in cognitive function during the first year. SOFT-EST evaluates estrogen levels during the first 4y of GnRH analogue and whether there is a suboptimally estrogen-suppressed subgroup. North American Pharmacogenetics study investigates whether genetic variations that affect T and E metabolism influence efficacy. Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr OT2-2-01.