Revisions in Implant-Based Breast Reconstruction: How Does Direct-to-Implant Measure Up?

Background: Immediate direct-to-implant breast reconstruction is increasingly performed for breast cancer treatment or prevention. The advantage over traditional tissue expander/implant reconstruction includes the potential for fewer surgical procedures. Methods: Retrospective, single-institution, three-surgeon review identified patients undergoing implant-based reconstruction from 2006 to 2011. Results: Six hundred eighty-two reconstructions were performed in 432 women with an average follow-up of 5 years. Four hundred sixty-five were direct-to-implant reconstructions with acellular dermal matrix while 217 were tissue expander/implant reconstructions without acellular dermal matrix. The overall revision rate in direct-to-implant reconstruction was 20.9 percent. There was no difference in total revision rates between direct-to-implant and tissue expander reconstruction (20.9 percent versus 20.3 percent; p = 0.861). Subgroup analysis showed no difference in revision for malposition (3.4 percent versus 5.5 percent; p = 0.200), size change (6.7 percent versus 5.5 percent; p = 0.569), fat grafting (8.6 percent versus 9.7 percent; p = 0.647), or capsular contracture (4.5 percent versus 3.2 percent; p = 0.429). Multivariable logistic regression analysis showed complications were associated with higher rates of revision for malposition or size in both groups (OR, 2.8; 95 percent CI, 1.56 to 5.13; p = 0.001). Smoking, preoperative irradiation, skin necrosis, and one surgeon were associated with higher rates of fat grafting, whereas increasing body mass index was associated with lower rates. Postoperative radiotherapy and hematoma were predictive of revision for capsular contracture. Conclusions: The 5-year revision rate in this series of direct-to-implant reconstruction was approximately 21 percent and similar to the revision rate in tissue expander/implant reconstruction. Surgical complications, radiotherapy, and the surgeon influenced the rate of revision similarly in both groups.