Breast cancer is the most common cancer in women worldwide, and is a leading cause of cancer death among women. Prophylactic or curative mastectomy is often followed by breast reconstruction for which there are several surgical approaches that use breast implants with which surgeons can restore the natural feel, size and shape of the breast.
To assess the effects of different types of breast implants on capsular contracture, surgical short- and long-term complications, postoperative satisfaction level and quality of life in women who have undergone reconstructive breast surgery after mastectomy.
We searched the Cochrane Breast Cancer Group's Specialised Register on 20 July 2015, MEDLINE (1985 to 20 July 2015), EMBASE (1985 to 20 July 2015) and the Cochrane Central Register of Controlled Trials (CENTRAL; Issue 8, 2015). We also searched the World Health Organization's International Clinical Trials Registry Platform (WHO ICTRP) and ClinicalTrials.gov on 16 July 2015.
We included randomised controlled trials (RCTs) and quasi-RCTs that compared different types of breast implants for reconstructive surgery. We considered the following types of intervention: implant envelope surfaces - texturised versus smooth; implant filler material - silicone versus saline, PVP-Hydrogel versus saline; implant shape - anatomical versus round; implant volume - variable versus fixed; brands - different implant manufacturing companies and implant generation (fifth versus previous generations).
Two review authors independently assessed methodological quality and extracted data. We used standard Cochrane methodological procedures. The quality of the evidence was assessed using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) system.
Five RCTs with 202 participants met the inclusion criteria. The women participants were typically in their 50s, and the majority of them (about 82%) received reconstructive surgery following breast cancer, while the others had reconstructive surgery after prophylactic mastectomy. The studies were heterogenous in terms of implant comparisons, which prevented us from pooling the data.The studies were judged as being at an unclear risk of bias for most risk of bias items owing to poor quality of reporting in the trial publications. Three of the five RCTs were judged to be at high risk of attrition bias, and one at high risk of detection bias.Textured silicone versus smooth silicone implants: textured implants were associated with worse outcomes when compared to smooth implants (capsular contracture: risk ratio (RR) 0.82, 95% CI 0.14 to 4.71; 1 study, 20 participants; very low quality evidence; reintervention: RR 0.82, 95% CI 0.14 to 4.71; 1 study, 20 participants; very low quality evidence). No results in this comparison were statistically significant.Silicone versus saline implants: saline-filled implants performed better than silicone-filled implants for some outcomes; specifically, they produced less severe capsular contracture (RR 3.25, 95% CI 1.24 to 8.51; 1 study, 60 participants; very low quality evidence) and increased patient satisfaction (RR 0.60, 95% CI 0.41 to 0.88; 1 study, 58 participants; very low quality evidence). However reintervention was significantly more frequent in the saline-filled implant group than in the silicone-filled group (OR 0.08, 95% CI 0.01 to 0.43; 1 study, 60 participants; very low quality evidence).Poly(N-vinyl-2-pyrrolidone) hydrogel-filled (PVP-hydrogel) versus saline-filled implants: PVP-hydrogel-filled implants were associated with worse outcomes when compared to saline-filled implants (capsular contracture: RR 3.50, 95% CI 0.83 to 14.83; 1 study, 40 participants; very low quality evidence; short-term complications: RR 2.10, 95% CI 0.21 to 21.39; 1 study, 41 participants; very low quality evidence).Anatomical versus round implants: anatomical implants were associated with worse outcomes than round implants (capsular contracture: RR 2.00, 95% CI 0.20 to 20.15; 1 study, 36 participants; very low quality evidence; short-term complications: RR 2.00, 95% CI 0.42 to 9.58; 1 study, 36 participants; very low quality evidence; reintervention: RR 1.50, 95% CI 0.51 to 4.43; 1 study, 36 participants; very low quality evidence). No results in this comparison were statistically significant.Variable-volume versus fixed-volume implants: data about one-stage reconstruction using variable-volume implants were compared with data about fixed-volume implants positioned during the second surgical procedure of two-stage reconstructions. Fixed-volume implant reconstructions were possibly associated with a greater number of women reporting that their reconstruction corresponded with expected results (RR 0.25, 95% CI 0.10 to 0.62; 1 study, 40 participants; very low quality evidence) and fewer reinterventions (RR 7.00, 95% CI 1.82 to 26.89; 1 study, 40 participants; very low quality evidence) when compared to variable-volume implants. A higher patient satisfaction level (rated from 1 to 6, with 1 being very bad and 6 being very good) was found with the fixed-volume implants for overall aesthetic result (mean difference (MD) -1.10, 95% CI -1.59 to -0.61; 1 study, 40 participants; very low quality evidence).There were no studies that examined the effects of recent (fifth) generation silicone implants versus previous generations or different implant manufacturing companies.
Despite the central role of breast reconstruction in women with breast cancer, the best implants to use in reconstructive surgery have been studied rarely in the context of RCTs. Furthermore the quality of these studies and the overall evidence they provide is largely unsatisfactory. Some of our results can be interpreted as early evidence of potentially large differences between different surgical approaches, which should be confirmed in new high-quality RCTs that include a larger number of women. These days - even after a few million women have had breasts reconstructed - surgeons cannot inform women about the risks and complications of different implant-based breast reconstructive options on the basis of results derived from RCTs.