The role of breast magnetic resonance imaging (MRI) in women newly diagnosed with breast cancer (BC) is controversial. This meta-analysis examines the effect of preoperative MRI compared with standard preoperative assessment on surgical outcomes, focusing on studies that used a controlled design.
Using random-effects logistic meta-regression modeling, we estimated the proportion of women with each outcome in the MRI versus no-MRI groups, and calculated the odds ratio (OR) and adjusted OR (adjusted for study-level median age, and, where appropriate, for temporal effect) for each model.
There were 9 eligible studies (2 randomized trials; 7 comparative cohorts). Outcomes in 3112 patients with BC (any histological tumor type) for MRI versus no-MRI (referent) were as follows: initial mastectomy 16.4% versus 8.1% [OR, 2.22 (P < 0.001); adjusted OR, 3.06 (P < 0.001)]; re-excision after initial breast conservation 11.6% versus 11.4% [OR, 1.02 (P = 0.87); adjusted OR, 0.95 (P = 0.71)]; overall mastectomy 25.5% versus 18.2% [OR, 1.54 (P < 0.001); adjusted OR, 1.51 (P < 0.001)]. In 766 patients with invasive lobular cancer (ILC), outcomes were as follows: initial mastectomy 31.1% versus 24.9% [OR, 1.36 (P = 0.056); adjusted OR, 2.12 (P = 0.008)]; re-excision after initial breast conservation 10.9% versus 18.0% [OR, 0.56 (P = 0.031); adjusted OR, 0.56 (P = 0.09)]; overall mastectomy 43.0% versus 40.2% [OR, 1.12 (P = 0.45); adjusted OR, 1.64 (P = 0.034)].
Our summary of the evidence showed that MRI significantly increased mastectomy rates and suggests an unfavorable harm-benefit ratio for routine use of preoperative MRI in BC. We found weak evidence that MRI reduced re-excision surgery in patients with ILC —although this was at the expense of increased mastectomies—and overall patient benefit from MRI in ILC is not clear from this study.
Supplemental Digital Content is available in the text.Meta-regression modeling of studies using a controlled design to examine the effect of preoperative magnetic resonance imaging (MRI) in breast cancer (3112 patients) showed that using MRI versus standard preoperative assessment (no-MRI) increased the overall mastectomy rate [25.5% vs 18.2%; odds ratio (OR), 1.54 (P < 0.001); adjusted OR, 1.51 (P < 0.001)] and did not reduce re-excision surgery in those with initial breast conservation [11.6% vs 11.4%; OR, 1.02 (P = 0.87); adjusted OR, 0.95 (P = 0.71)].
*Screening and Test Evaluation Program, School of Public Health, Sydney Medical School, University of Sydney, Sydney, Australia
†Breast Service, Department of Surgery, Memorial Sloan-Kettering Cancer Center, New York, NY.
Reprints: Nehmat Houssami, MBBS, PhD, School of Public Health (A27), Sydney Medical School, University of Sydney, Sydney 2006, Australia. E-mail: firstname.lastname@example.org.
Disclosure: Supported by National Health and Medical Research Council program grant No. 633003 to the Screening and Test Evaluation Program.
Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal's Web site (www.annalsofsurgery.com).