Send to

Choose Destination

See 1 citation in J Clin Oncol 2014:

J Clin Oncol. 2014 Feb 10;32(5):392-401. doi: 10.1200/JCO.2013.52.7515. Epub 2014 Jan 6.

An individual person data meta-analysis of preoperative magnetic resonance imaging and breast cancer recurrence.

Author information

Nehmat Houssami, Robin Turner, Petra Macaskill, the Sydney Medical School, University of Sydney, Sydney, New South Wales, Australia; Lindsay W. Turnbull, Centre for Magnetic Resonance Investigations, Hull York Medical School in association with University of Hull, Hull, United Kingdom; David R. McCready, Princess Margaret Cancer Centre, University of Toronto, Toronto, Ontario, Canada; Todd M. Tuttle, University of Minnesota, Minneapolis, MN; Neha Vapiwala and Lawrence J. Solin, University of Pennsylvania School of Medicine; Lawrence J. Solin, Albert Einstein Medical Center, Philadelphia, PA.



There is little consensus regarding preoperative magnetic resonance imaging (MRI) in breast cancer (BC). We examined the association between preoperative MRI and local recurrence (LR) as primary outcome, as well as distant recurrence (DR), in patients with BC.


An individual person data (IPD) meta-analysis, based on preoperative MRI studies that met predefined eligibility criteria, was performed. Survival analysis (Cox proportional hazards modeling) was used to investigate time to recurrence and to estimate the hazard ratio (HR) for MRI. We modeled the univariable association between LR (or DR) and MRI, and covariates, and fitted multivariable models to estimate adjusted HRs. Sensitivity analysis was based on women who had breast conservation with radiotherapy.


Four eligible studies contributed IPD on 3,180 affected breasts in 3,169 subjects (median age, 56.2 years). Eight-year LR-free survival did not differ between the MRI (97%) and no-MRI (95%) goups (P = .87), and the multivariable model showed no significant effect of MRI on LR-free survival: HR for MRI (versus no-MRI) was 0.88 (95% CI, 0.52 to 1.51; P = .65); age, margin status, and tumor grade were associated with LR-free survival (all P < .05). HR for MRI was 0.96 (95% CI, 0.52 to 1.77; P = .90) in sensitivity analysis. Eight-year DR-free survival did not differ between the MRI (89%) and no-MRI (93%) groups (P = .37), and the multivariable model showed no significant effect of MRI on DR-free survival: HR for MRI (v no-MRI) was 1.18 (95% CI, 0.76 to 2.27; P = .48) or 1.31 (95% CI, 0.76 to 2.27; P = .34) in sensitivity analysis.


Preoperative MRI for staging the cancerous breast does not reduce the risk of LR or DR.

[Indexed for MEDLINE]

Supplemental Content

Full text links

Icon for Atypon
Loading ...
Support Center