SOFT/TEXT Final 15-Year Data: Ovarian Suppression Plus Exemestane Cuts Recurrence in Premenopausal Breast Cancer

Why This Study Matters

Breast cancer caused an estimated 670,000 deaths worldwide in 2022 among 2.3 million new diagnoses, and premenopausal women in low- and middle-income countries carry a disproportionate share of that burden. Hormone receptor (HR)-positive disease is generally considered a favorable subtype, but it carries an elevated risk of late distant recurrence — a risk that is more pronounced in women diagnosed before age 35.

The companion Suppression of Ovarian Function Trial (SOFT) and Tamoxifen and Exemestane Trial (TEXT) were designed two decades ago to answer whether adding ovarian function suppression (OFS) to adjuvant endocrine therapy — and pairing OFS with an aromatase inhibitor rather than tamoxifen — meaningfully reduces recurrence and death in premenopausal women with HR-positive early breast cancer. This report presents the trials' final, mature 15-to-17-year outcomes, the longest follow-up yet published for this patient population.

Study Design

SOFT (n=3,047 intention-to-treat; enrolled December 2003–January 2011) randomized premenopausal women with resected, HR-positive breast cancer 1:1:1 to tamoxifen 20 mg daily for 5 years; tamoxifen plus OFS for 5 years; or exemestane 25 mg daily plus OFS for 5 years. OFS was delivered by triptorelin 3.75 mg intramuscularly every 4 weeks, bilateral oophorectomy, or ovarian irradiation, with 91% of OFS patients initially choosing triptorelin. Randomization was stratified by prior chemotherapy, nodal status, and intended OFS method.

TEXT (n=2,660 intention-to-treat; enrolled November 2003–April 2011) randomized patients 1:1 to exemestane plus triptorelin or tamoxifen plus triptorelin for 5 years, with oophorectomy or ovarian irradiation permitted after 6 months of triptorelin. Randomization was stratified by intended chemotherapy use and nodal status.

The primary endpoint in both trials was disease-free survival (DFS) — time to invasive breast cancer recurrence (local, regional, or distant), invasive contralateral breast cancer, a second non-breast invasive cancer, or death. Secondary endpoints included breast cancer-free interval (BCFI), distant recurrence-free interval (DRFI), and overall survival (OS). The combined SOFT+TEXT analysis of exemestane plus OFS versus tamoxifen plus OFS included 4,690 women in the intention-to-treat population.

Patient Population

Before randomization in SOFT, 53.4% of women (1,628) had received neoadjuvant or adjuvant chemotherapy; 34.5% had node-positive disease and 84.9% had HER2-negative tumors. In TEXT, 60.4% (1,607) received chemotherapy after randomization. Across the combined population, 42.2% had node-positive tumors and 86.0% had HER2-negative disease. More than two-thirds of surviving patients in each trial contributed follow-up data during 2023–2024.

Primary Endpoint and Efficacy Results

P-values are reported here only where the source explicitly states them; cells without a P-value reflect estimates for which the study did not report a formal significance test alongside the hazard ratio.

The breast cancer-free interval comparison in the combined analysis was reported as 946 events, a hazard ratio, and a P-value, without arm-specific 15-year percentages in the source publication.

Subgroup Analyses

The study authors describe subgroup analyses of DFS and BCFI in SOFT as showing treatment-effect estimates consistent with earlier analyses, with persisting evidence of treatment-effect heterogeneity by HER2 status — a qualitative statement in the source rather than a quantified subgroup-interaction test. Several of the subgroup comparisons above, including the age-under-35 and chemotherapy-omitted cohorts, are descriptive and were not accompanied by a hazard ratio or confidence interval in the published report.

Safety Profile

Late adverse events were assessed among patients who remained event-free at 6 years, tracking cumulative incidence over the following 9 years.

No excess of deaths without a prior breast cancer event occurred among patients assigned OFS, and acute myeloid leukemia was rare overall — one case among the SOFT prior-chemotherapy cohort (0.06%).

Interpretation and Broader Context

The final 15-year data confirm that adding OFS to oral endocrine therapy reduces recurrence in premenopausal HR-positive breast cancer, with the largest benefit when OFS is paired with an aromatase inhibitor. The authors are explicit, however, that this recurrence benefit did not uniformly translate into a survival benefit for the trial population as a whole.

Most premenopausal women will not derive a long-term overall-survival advantage from intensification of endocrine therapy, but some — particularly younger premenopausal women, and those with grade 3, HER2-negative tumors — can derive meaningful improvements in their risk of distant metastasis or death.
— Discussion, Francis et al., Annals of Oncology, 2026

In the SOFT no-chemotherapy cohort — predominantly older, lower-risk patients with small, lower-grade tumors — 15-year DRFI and OS remained high (DRFI above 94%) regardless of which endocrine therapy patients received, even though BCFI events roughly doubled between the 8-year and 15-year analyses. The authors frame this as evidence that nuanced, risk-based counseling, rather than blanket escalation of therapy, is warranted for lower-risk premenopausal patients, while intensified therapy is more clearly justified in very young patients and those with high-grade, HER2-negative disease.

The paper also cites external context for interpreting its own findings: a meta-analysis reporting a significant 10-year breast-cancer-mortality reduction with OFS (relative risk 0.74, 95% CI 0.58–0.95, P=0.012) in women under 45 receiving tamoxifen without chemotherapy, and the ASTRRA trial's 10-year outcomes. These are referenced by the SOFT/TEXT investigators as supporting evidence from other trials, not as within-trial results of SOFT or TEXT.

Future Directions

For very young, high-risk premenopausal patients who continue to face substantial recurrence risk even on exemestane plus OFS, the authors point to strategies under active investigation: baseline breast MRI, germline genetic testing (with consideration of bilateral mastectomy and adjuvant olaparib in the highest-risk carriers), improved endocrine-therapy adherence, extended endocrine therapy beyond 5 years, adjuvant CDK4/6 inhibitors, and combining immunotherapy with neoadjuvant chemotherapy in high-grade disease.

Limitations

SOFT and TEXT were open-label rather than blinded, which leaves the possibility of assessment bias for any outcome with a subjective component, though DFS and OS are driven largely by objective events. Several subgroup findings emphasized in this report — the age-under-35 overall-survival comparison, the grade 3 HER2-negative absolute-difference figures, and the node-positive, chemotherapy-omitted cohort — are descriptive, non-randomized comparisons reported without a hazard ratio or confidence interval, and should be read as hypothesis-generating rather than confirmatory given their small size. Despite reaching statistical significance for disease-free survival and distant recurrence, exemestane plus OFS did not produce a significant overall-survival benefit over tamoxifen plus OFS in either trial's intention-to-treat population at 15 years, so this analysis cannot establish that intensified endocrine therapy extends life for the average enrolled patient — only that it reduces recurrence, most clearly in the youngest and highest-risk women. The combined SOFT+TEXT analysis also pools two trials with different randomization designs (a three-arm comparison in SOFT versus a two-arm comparison in TEXT), a planned pooled analysis rather than a single randomized comparison. Finally, the late fracture data cover only the 9 years following year 6 of follow-up and do not report a combined-analysis fracture rate, leaving the longer-term skeletal cost of extended OFS incompletely characterized in this source.