NeoSphere

Design, population and eligibility

Treatment and comparator

Four neoadjuvant cycles: (A) trastuzumab 8 mg/kg loading then 6 mg/kg q3w plus docetaxel 75 mg/m² q3w (escalated to 100 mg/m² if tolerated); (B) pertuzumab 840 mg loading then 420 mg q3w plus trastuzumab/docetaxel; (C) pertuzumab plus trastuzumab; (D) pertuzumab plus docetaxel. Surgery followed, with protocol-directed postoperative chemotherapy and HER2-directed therapy.

Primary and secondary endpoints

Primary endpoint: pathological complete response in the breast (ypT0/is; primary publication definition). Secondary endpoints included pCR in breast and nodes, safety, 5-year PFS and DFS.

Key efficacy results

Breast pCR was 45.8% (49/107) with pertuzumab/trastuzumab/docetaxel versus 29.0% (31/107) with trastuzumab/docetaxel; absolute difference 16.8 percentage points (95% CI 3.5–30.1; P=0.0141). At 5 years, PFS was 86% (95% CI 77–91) versus 81% (71–87); HR 0.69 (95% CI 0.34–1.40). Five-year DFS was 84% versus 81%. The survival comparison was descriptive and imprecise.

Source-reported outcome visualization

Breast pCR (primary endpoint) and five-year PFS.

pCR is the primary response endpoint. Five-year PFS is shown as descriptive follow-up; confidence interval crosses 1 and does not establish survival superiority.

Absolute effect

Breast pCR increased by 16.8 percentage points (45.8% vs 29.0%); approximate NNT 6 neoadjuvant patients treated with dual blockade plus docetaxel for one additional breast pCR. This is an NNT for pCR, not cure or survival. The 5-year PFS difference was 5 points, but the HR CI was wide (0.34–1.40); no survival NNT is assigned.

Trial-specific toxicity and treatment burden

Any treatment-emergent AE occurred in approximately 98% in the dual-blockade/docetaxel arm and 100% in the trastuzumab/docetaxel arm. Grade ≥3 AEs were common, predominantly neutropenia (44.9% vs 57.0% in the regulatory evidence synthesis). Diarrhoea was more frequent with pertuzumab. Treatment involved four neoadjuvant cycles plus postoperative treatment; the small phase II trial is not powered for rare cardiotoxicity.

Quality of life and patient-reported outcomes

No robust randomized QoL benefit was established as a primary outcome. Treatment burden and chemotherapy toxicity are substantial in both docetaxel-containing arms.

Subgroups and applicability

pCR appeared higher in hormone-receptor-negative than hormone-receptor-positive tumours. The trial was not powered for definitive subgroup treatment interactions. The pCR definition used in the primary endpoint was breast-only (ypT0/is), not necessarily total pCR ypT0/is ypN0.

Current-practice context · reviewed 3 November 2026

NeoSphere established proof of concept for neoadjuvant dual HER2 blockade. Contemporary regimens and post-neoadjuvant escalation have evolved; use the trial as foundational biologic and response evidence rather than as a direct comparison of current full perioperative strategies.

Limitations and critical interpretation

Phase II, four-arm study with modest sample sizes; primary endpoint was pCR, not survival. 5-year PFS HR confidence interval crosses 1.0. Postoperative therapy was protocol-directed and subsequent standards have changed.

Bottom line: NeoSphere demonstrated a significant pCR gain from adding pertuzumab to trastuzumab/docetaxel. It established the neoadjuvant dual-blockade concept, but does not independently prove an OS advantage.

Primary sources and follow-up publications

1. Primary report, Lancet Oncology 2012
2. Five-year follow-up, Lancet Oncology 2016
3. ClinicalTrials.gov

Outcome values are sourced from the linked reports. Unreported values are not imputed. NNT/NNH is only stated when an explicit fixed-time absolute difference supports a defensible calculation; no NNT is derived from median survival.