Adjuvant Capecitabine Fails to Provide Survival Benefit in Early-Stage TNBC


Adding adjuvant capecitabine to the standard treatment for patients with early-stage triple-negative breast cancer did not lead to a significant improvement in disease-free or overall survival compared with observation, according to findings presented during the 2018 San Antonio Breast Cancer Symposium

Miguel Martin, MD, PhD

Adding adjuvant capecitabine to the standard treatment for patients with early-stage triple-negative breast cancer (TNBC) did not lead to a significant improvement in disease-free (DFS) or overall survival (OS) compared with observation, according to findings presented during the 2018 San Antonio Breast Cancer Symposium.1

“We were disappointed to find that adding adjuvant capecitabine to the standard treatment did not significantly improve disease-free or overall survival,” said Miguel Martín, MD, PhD, lead author of the phase III GEICAM/CIBOMA trial and a professor of medicine and head of the Medical Oncology Service at Hospital Gregorio Marañón, Universidad Complutense, Madrid, Spain.

Additionally, data from a subset analysis showed that patients with nonbasal-like TNBC who received capecitabine had a 49% reduction in the risk of disease progression and a 52% reduction in the risk of death versus those with nonbasal-like disease on the observation arm.

“However, given that we found a subset of the patients with nonbasal-like disease seemed to have a significant benefit from capecitabine, and data from the CREATE-X trial showed that adjuvant capecitabine significantly reduced the rate of relapse and improved overall survival when administered to breast cancer patients with residual disease after neoadjuvant chemotherapy, we strongly recommend that patients with triple-negative breast cancer discuss adjuvant capecitabine with their oncologists.”

Findings from the CREATE-X trial previously showed that the addition of adjuvant capecitabine following standard neoadjuvant chemotherapy containing anthracycline, taxane, or both was safe and effective in prolonging DFS and OS among patients with HER2-negative breast cancer who had residual invasive disease on pathological testing.2

“Patients with early-stage triple-negative breast cancer are usually treated with surgery and chemotherapy, and sometimes radiotherapy,” concluded Martín. “New therapeutic approaches are urgently needed, however, because the risk of relapse is high: 7% to 10% of those with stage I disease relapse, 15% to 20% of those with stage II disease, and 25% to 50% of those with stage III disease.”

In the randomized, phase III GEICAM/CIBOMA trial, 876 patients with early-stage TNBC that was estrogen receptor—, progesterone receptor–, and HER2-negative who underwent surgery and chemotherapy were randomized 1:1 to receive 8 cycles of oral capecitabine at 1000 mg/m2twice daily for 14 days every 3 weeks (n = 448) or observation (n = 428).

Patients had stage T1c-3, N0-N3a, or M0 disease. To be eligible for enrollment, patients also had to have received standard neoadjuvant and adjuvant chemotherapy with anthracycline with or without taxanes, and underwent surgery with negative margins. Six cycles of chemotherapy were mandatory except for patients with N0 tumors.

Patients were stratified by institution, basal phenotype according to CK5/6 and/or EGFR staining, number of axillary lymph nodes (0 vs 1-3 vs ≥4), and prior chemotherapy with anthracycline versus anthracycline plus taxanes. The primary endpoint was DFS and OS, subgroup analyses, safety, and biomarkers.

The median age was 49.5 years, with 60.7% of patients from Spain and 39.3% from Latin America. Most patients were postmenopausal (68.4%) compared with premenopausal (31.5%). The majority of patients were stage II in both arms (60.3% with capecitabine vs 63.3% with observation). In the capecitabine arm, 54.5% of patients had node-negative tumors, 27% with 1 to 3 positive nodes, 17.2% with ≥4 positive nodes, and 1.3% had unavailable status. Among those on the observation arm, 56.5% had node-negative status, 29.0% had 1 to 3 positive nodes, 14.3% had ≥4 positive nodes, and 0.2% had unavailable node status.

Adjuvant chemotherapy alone was given in 78.8% of those on the capecitabine arm and 82.2% on observation, while neoadjuvant and adjuvant chemotherapy was given in 19.9% and 17.5% in capecitabine and observation, respectively. These data were unavailable for 1.3% and 0.2% of the capecitabine and observation groups, respectively.

Pathologic complete response (pCR) to neoadjuvant chemotherapy among arms was comparable (24.7% on capecitabine vs 25.3% on observation).

At a median follow-up of 7.3 years, results showed that the 5-year DFS rates were 79.6% and 76.8% in the capecitabine and observation arms, respectively, which was not deemed statistically significant (P= .135). There were 105 events in the capecitabine arm and 120 in the observation arm (HR, 0.82; 95% CI, 0.63-1.06;P= .135). The adjusted hazard ratio for stratification variables was 0.79 (95% CI, 0.61-1.03;P= .082).

“There was a nonsignificant trend in favor of capecitabine, but the trial had only 876 participants, which means it was not statistically powered to identify small but clinically relevant differences,” noted Martín. “One possible reason for the discrepancy in the results of the CREATE-X trial and our trial may be that the populations had different prognostic features; the risk of relapse of our population was much less than in the CREATE-X trial.”

The 5-year OS rates were 86.2% and 85.9% with capecitabine and observation, respectively (P= .623).

The main limitation of the study, Martín said, was its limited power to show small but clinically relevant improvements in outcome with capecitabine; this is due to the sample size and the lower than expected number of relapse events in the control arm.

DFS data of a subgroup analysis of patients with nonbasal-like disease was reported. In the basal-like cohort, the 5-year DFS rates were 78.5% with capecitabine and 78.2% on observation (HR, 0.94; 95% CI, 0.70-1.27;P= .696). For patients with nonbasal-like disease treated with capecitabine, the 5-year DFS rate was 82.6% and 72.9% with observation (HR, 0.53; 95% CI, 0.31-0.91;P= .02).

In the basal-like disease population, the 5-year OS rates were 84.9% with capecitabine and 88.0% with observation (P= .286). For those with nonbasal-like disease, these rates were 89.5% with capecitabine and 79.6% with observation (P= .007).

“This is an intriguing finding,” said Martín. “However, it should be interpreted with caution because the interaction test was negative for disease-free survival (P= .0694), although it was statistically significant for overall survival (P= .0052).”

Additionally, in the overall population, there were 14 deaths on the capecitabine arm versus 10 on the observation arm. Four patients on observation experienced locoregional recurrence, and 0 did on the capecitabine arm; 5 patients on capecitabine and 12 on observation had ipsilateral breast cancer recurrence, while 12 and 14 had contralateral invasive breast cancer. The rates of distant recurrence were similar between arms at 14.3% and 15.4% with capecitabine and observation, respectively. However, in the nonbasal-like population, the distant recurrence rates were 10.9% with capecitabine and 16.4% with observation.

Regarding safety, tolerance of extended adjuvant capecitabine was as expected, with a median dose intensity of 86.3% and 75.2% of patients who completed the planned 8 cycles.


  1. Martín M, Barrios CH, Torrecillas L, et al. Efficacy results from GEICAM/2003-11_CIBOMA/2004-01 study: a randomized phase III trial assessing adjuvant capecitabine after standard chemotherapy for patients with early triple negative breast cancer. In: Proceedings from the 2018 San Antonio Breast Cancer Symposium; December 4-8, 2018; San Antonio, Texas. Abstract LBA_2.
  2. Masuda N, Lee S-J, Ohtani S, et al. Adjuvant capecitabine for breast cancer after preoperative chemotherapy.N Eng J Med. 2017;376:2147-2159. doi: 10.1056/NEJMoa1612645.
Related Videos
Video 5 - "Second-Line Treatment Considerations and Improving Outcomes in Breast Cancer"
Video 4 - "Impressions on Safety Data from EMERALD Subgroup Analysis"
Related Content