ESMO 2014: Bevacizumab Maintenance in HER2-Negative Metastatic Breast Cancer

Efficacy and safety of bevacizumab in the randomised phase III IMELDA and TANIA clinical trials

During the Proffered Paper session in Metastatic Breast Cancer at ESMO Congress 2014 in Madrid, Spain, the researchers presented results of IMELDA and TANIA randomised phase III clinical trials in HER2-negative metastatic breast cancer. In the IMELDA study, adding capecitabine to maintenance bevacizumab provided statistically significant and clinically meaningful improvements in progression-free survival (PFS) and overall survival (OS). The primary objective of TANIA was also met, showing statistically significantly improvement in PFS with bevacizumab after progression on first-line bevacizumab-containing therapy.

IMELDA: Efficacy and safety of maintenance bevacizumab with or without capecitabine after initial first-line bevacizumab plus docetaxel

Prolonging first-line chemotherapy results with maintenance treatment may influence OS. In HER2-negative patients with locally recurrent/metastatic breast cancer, combining bevacizumab with first-line chemotherapy significantly improves PFS. Bevacizumab benefit is most pronounced when combined with a taxane. Cumulative toxicity prevents taxane continuation until disease progression.

Until regulatory withdrawal of bevacizumab/docetaxel in 2011, this combination was considered as a valid first-line option for HER2-negative metastaic breast cancer based on results of a phase III trial. The PFS and response rate (RR) with maximum 9 cycles of first-line docetaxel were significantly improved by adding bevacizumab continued until disease progression.

The open-label randomised phase III IMELDA trial tested whether switching to a more tolerable chemotherapy with a different mechanism of action while continuing VEGF inhibition may be more effective. It was meant that by adding capecitabine to maintenance bevacizumab continued until disease progression after initial bevacizumab/docetaxel will improve PFS. The study findings were presented by Dr Joseph Gligorov of the Medical Oncology Department, Tenon University Hospital, Paris, France.

Patients with HER2-negative measurable metastatic breast cancer, ECOG PS <2 and no prior chemotherapy were eligible for study inclusion.

After 3‒6 cycles of bevacizumab/docetaxel, patients without disease progression were randomised to bevacizumab alone or bevacizumab/capecitabine until disease progression. Stratification factors were oestrogen receptor (ER) status, presence of visceral metastases, response status and LDH concentration.

ESMO Congress 2014 352O IMELDA PFS

PFS from time of randomisation. © Joseph Gligorov

The primary endpoint was PFS from randomisation to progression/death; secondary endpoints included RR, clinical benefit rate, time to disease progression, OS from randomisation, safety and quality of life. The sample size was calculated based on a PFS hazard ratio (HR) of 0.70 with median PFS improvement from 5.8 to 8.3 months. In total 360 enrolled patients were required for 290 randomised patients. It was planned that 244 PFS events provide 80% power at 5% 2-sided alpha. The study was not designed for formal OS comparison.

Between June 2009 and March 2011 when enrolment was prematurely terminated, 287 patients were enrolled and 284 of them treated with 185 (65%) who completed initial treatment and subsequently randomised to maintenance treatment. The pprotocol was amended to continue follow-up for 2 years after last randomisation.

Median age in the bevacizumab arm was 54 years and 49 years in the bevacizumab/capecitabine arm. Triple-negative disease was recorded in 22% of patients included in the bevacizumab arm and 27% of patients in the bevacizumab/capecitabine arm. Visceral metastases were nearly identical in both group (69% vs. 68%), however their presence in ≥3 organs was higher in bevacizumab arm at enrolment to the initial phase (57% vs. 47%).

Results of the IMELDA Study

ESMO Congress 2014 352O IMELDA OS

OS from time of randomisation. © Joseph Gligorov

In the maintenance arm, median treatment duration was longer in bevacizumab/capecitabine group (8.3 vs 3.5 months). Adding capecitabine to maintenance bevacizumab provided statistically significant and clinically meaningful improvements in PFS from time of randomisation (HR 0.38, p<0.001; median 11.9 vs 4.3 months) and exploratory analysis (PFS from start of first-line therapy), as well as improvement in median OS from time of randomisation (HR 0.42, p<0.001; 39 vs 23.3 months), despite the smaller than planned sample size because of early termination of accrual. However, at the time of report there was insufficient duration of OS follow-up with low event rate.

There was a manageable increase in adverse events mainly due to hand-foot syndrome experienced in 33% of patients in bevacizumab/capecitabine arm. Hypertension was recorded in 9% of patients in the combined arm and 3% of patients in the bevacizumab only arm. The rate of proteinuria was same (4%) in both groups. Gastroenteritis occured in 3 patients in the bevacizumab single agent arm.

Conclusions of the IMELDA Study

Dr Gligorov concluded that in patients benefiting from first-line bevacizumab-containing therapy, continued bevacizuamb with capecitabine improves efficacy. Ongoing evaluation considers collection of data on anti-cancer treatment after study therapy and patient-reported outcomes.

Dr Hope Rugo, who discussed the study results, said that the IMELDA trial tried to answer the question: does continuing bevacizumab post progression have an impact on outcome? Dr Rugo questioned if bevacizumab is an adequate maintenance therapy after response to first-line chemotherapy. In the study, there was a longer duration of treatment with capecitabine/bevacizumab vs bevacizumab (twice the number of cycles), markedly longer PFS, and the PFS from start of first-line was doubled, OS was significantly longer, but there were also almost double the grade > 3 toxicity (mainly hand-foot syndrome, hypertension, but also thromboembolism). Hand-foot syndrome resulted in discontinuation of capecitabine in 10% of patients.

TANIA: Efficacy and safety of continued or reintroduced bevacizumab after first-line bevacizumab for HER2-negative locally recurrent/metastatic breast cancer

Combining bevacizumab with first- or second-line chemotherapy in randomised phase III trials showed significantly improved PFS in HER2-negative locally recurrent/metastatic breast cancer. Sustained VEGF blockade may be important for long-term disease control. The open-label randomised phase III TANIA study evaluated further bevacizumab in bevacizumab-pretreated locally recurrent/metastatic breast cancer. The study results were presented by Gunther von Minckwitz, Managing Director of the German Breast Group and University Women's Hospital, Neu-Isenburg, Germany.

Patients with HER2-negative locally recurrent/metastatic breast cancer who had progressed during/after ≥12 weeks of first-line bevacizumab plus chemotherapy were randomised 1:1 to second-line single-agent chemotherapy either alone or with bevacizumab (15 mg/kg q3w or 10 mg/kg q2w).

Stratification factors were: hormone receptor status; time to first-line progression (<6 vs ≥6 months); chemotherapy choice (taxane vs non-taxane vs vinorelbine); and LDH concentration (≤1.5 vs >1.5 × upper normal limit).

Second-line therapy was continued until disease progression, unacceptable toxicity or consent withdrawal. At disease progression, patients in the chemotherapy arm received thirrd-line chemotherapy without bevacizumab (no crossover); patients initially randomised to chemotherapy plus bevacizumab received third-line chemotherapy plus bevacizumab.

Chemotherapy options were based on investigator’s choice, but doublets were not allowed: paclitaxel, nab-paclitaxel, docetaxel, capecitabine, gemcitabine, pegylated liposomal doxorubicin, non-pegylated liposomal doxorubicin, doxorubicin, epirubicin, vinorelbine, cyclophosphamide, ixabepilone and in 3rd line only eribulin.

The primary endpoint was PFS from randomisation to second disease progression/death. Additional endpoints included second-line PFS in prespecified subgroups, second- and third-line PFS calculated from randomisation to third disease progression/death, second-line objective response rate, OS, safety, QoL and biomarkers.

Sample size was calculated based on assuming prolonging median PFS from 7 to 9.3 months and a HR of 0.75. PFS events were required in 384 of 488 patients for 80% power at 2-sided α=0.05.

Results of the TANIA Study

ESMO Congress 2014 353O TANIA PFS

Primary endpoint – Second-line PFS. © Gunter von Minckwitz

At ESMO 2014, the investigators presented the mature pre-specified second-line PFS analysis. Endpoints relating to third-line therapy will be presented at the final analysis.

From January 2011 to April 2013, 494 patients were enrolled (247 in chemotherapy arm and 247 in chemotherapy plus bevacizumab arm). Baseline characteristics were similar in chemotherapy vs chemotherapy plus bevacizumab groups: median age 54 vs 56 years; triple negative disease 23.1% vs 19.8%; disease-free interval ≤12 months 9.7% vs 7.3%.

The most frequently chosen second-line chemotherapy was capecitabine, 59.7% in chemotherapy group and 60.4% in the chemotherapy plus bevacizumab group.

Median follow-up was similar in both groups. At data cut-off on 20 December 2013, median second-line PFS was 4.2 months in chemotherapy vs 6.3 months in chemotherapy/bevacizumab groups (stratified HR = 0.75, p = 0.0068).  Subgroup analysis for PFS by stratification factor was also more favourable for the bevacizumab/chemotherapy grooup.

The best objective response ratewas not statistically different in two groups (16.8% vs 20.9%). However, the stable disease was recorded in 33.5% patients in the chemotherapy arm and 48.9% in the bevacizumab/chemotherapy arm.

Median duration of response was 10.6 vs 8.3 months for chemotherapy and bevacizumab/chemotherapy paients.

The rate of side effects was slighlty higher in the chemotherapy/bevacizumab arm:

  • hypertension (7.1% vs 13.5%),
  • proteinuria (0.4% vs 6.9%),
  • venous thromboembolic event (2.1% vs 3.3%),
  • febrile neutropaenia (1.7% vs 3.3%),
  • congestive heart failure (0.4% vs 2.0%),
  • bleeding (1.7% vs 0.4%),
  • arterial thromboembolic event (1.3% vs 0%),
  • wound-healing complication (0% vs 0.8%),
  • gastrointestinal perforation (0% vs 0.4%),
  • and fistula/abscess (0% in both groups).

Conclusions of the TANIA Study

The authors concluded that the primary objective of the study was met, showing statistically significantly improved PFS with bevacizumab after disease progression on first-line bevacizumab-containing therapy. Second-line safety results were as expected from previous bevacizumab trials in locally recurrent/metastatic breast cancer. Final OS, PFS from randomisation to third-line progression/death and third-line safety results are anticipated in mid 2015.

Dr Hope Rugo, who discussed the results from this study as well, said that the TANIA trial tried to see the impact of chemotherapy vs targeted therapy alone as maintenance therapy after response. Almost 85% of patients received first-line taxane (73% paclitaxel). There was an unusually long PFS in first-line. Almost 60% received second-line capecitabine. The PFS increased with continued bevacizumab; there was an increase in stable disease but not ORR and, as in prior studies, the greater benefit was in triple-negative breast cancer. Furthermore, there was more toxicity with bevacizumab (hypertension, proteinuria, neutropenia).

Maintenance chemotherapy improves PFS and OS after response to first-line chemotherapy. Unclear additional benefit from bevacizumab must be balanced against cost and toxicity. Bevacizumab alone should not be used as maintenance therapy in this setting. Almost 75% had hormone receptor-positive disease and there might be a role of maintenance hormone therapy.

Dr Rugo concluded that at present the role of bevacizumab is unclear in breast cancer.


Upon the presentation at ESMO 2014, the study results of both IMELDA and TANIA trials are published in the Lancet Oncology.


The IMELDA and TANIA trials were sponsored by F. Hoffmann-La Roche.