77P - Retrospective audit to assess impact of tumour biology on locoregional treatment outcome in breast cancer

Date 20 December 2015
Event ESMO Asia 2015 Congress
Session Poster presentation 2
Topics Breast Cancer
Pathology/Molecular Biology
Basic Scientific Principles
Presenter Pooja Padmanaban
Citation Annals of Oncology (2015) 26 (suppl_9): 16-33. 10.1093/annonc/mdv519
Authors P. Padmanaban1, V. Parmar1, N. Nair1, R.A. Badwe1, N. Hariharan1, R. Hawaldar2, V. Vanmali2, A. Bansode2, S. Siddique2
  • 1Breast Services, Surgical Oncology, Tata Memorial Hospital Centre, 400012 - Mumbai/IN
  • 2Institutional Review Board, Tata Memorial Hospital Centre, 400012 - Mumbai/IN



Extent of surgery in breast cancer is based on breast-tumor ratio and radiological features and not on biological tumour subtype. We wish to assess association of tumour biology and locoregional treatment.


A retrospective audit of 2240 women with unilateral stage I-III breast cancer treated at a single institution during 2009-2011. Based on IHC, 3 subtypes were identified namely, Hormone receptor positive (ER &/or PgR +ve & Her2neu negative), Her 2 neu positive (Her 2 neu score 3 + , or 2+ & FISH amplified, with any ER/PR), and Triple negative/TNBC(ER/PgR & Her 2 neu negative). Tumors reported Her 2 neu equivocal without confirmatory FISH result were excluded. Locoregional-recurrence free survival (LRFS) was plotted using Kaplan Meier graphs. Cox regression models were used to assess interaction with known prognostic factors. Objectives were to assess LRFS within subtypes depending on extent of locoregional treatment.


Median age was 48 yrs (range 18-84), 1613(66.1%) were operable breast cancers, 627(33.9%) were locally advanced; median cT was 4.3cm (range 1.5-18), 1047(46.7%) were HR + , 491(21.9%) Her 2 neu + and 702(31.3%) were TNBC. NACT was administered in 1413(63.1%), 827(36.9%) were operated upfront. 1208(53.9%) underwent BCT, 1032(46.1%) underwent MRM. At median follow up of 45mth, LRFS was significantly better for HR+ than TNBC and Her 2 neu+ for both BCT(p = 0.0001) and MRM(p = 0.005). However, no significant difference in LRFS was seen within each subtype with respect to extent of surgery (HR+ p = 0.125, TNBC p = 0.430, HER2neu+ p = 0.444). By Cox regression, younger age (HR 1.026 95%CI 1.009–1.043, p = 0.003), nodal positivity (HR 2.316, 95%CI 1.578–3.399, p = 0.0005) and advanced stage (HR 1.669, 95%CI 1.079–2.581, p = 0.021) were associated with worse LRFS. HR+ had best LRFS, followed by TNBC (HR 3.316, 95%CI 1.783–6.164, p = 0.0005), and Her 2 neu+ (HR 4.169, 95%CI 2.180–7.974, p = 0.0005). Extent of surgery had no impact on LRFS within each subtypes.


LRFS was dictated by disease biology rather than extent of local surgery. Within each subtype, type of surgery did not impact LRFS. Younger age, nodal positivity and advanced stage were associated with worse LRFS.

Clinical trial identification


All authors have declared no conflicts of interest.