First‐year complications after immediate breast reconstruction with a biological and a synthetic mesh in the same patient: A randomized controlled study

Abstract Background Even though meshes and matrices are widely used in breast reconstruction, there is little high‐quality scientific evidence for their risks and benefits. The aim of this study was to compare first‐year surgical complication rates in implant‐based immediate breast reconstruction with a biological mesh with that of a synthetic mesh, in the same patient. Methods This study is a clinical, randomized, prospective trial. Patients operated on with bilateral mastectomy and immediate breast reconstruction were randomized to biological mesh on one side and synthetic mesh on the other side. Results A total of 48 breasts were randomized. As the synthetically and the biologically reconstructed breasts that were compared belonged to the same woman, systemic factors were exactly the same in the two groups. The most common complication was seroma formation with a frequency of 38% in the biological group and 3.8% in the synthetical group (p = .011). A higher frequency of total implant loss could be seen in the biologic mesh group (8.5% vs. 2%), albeit not statistically significant (p = .083). Conclusions In the same patient, a synthetic mesh seems to yield a lower risk for serious complications, such as implant loss, than a biological mesh.

T A B L E 1 Randomized controlled trials on matrices and meshes with complications as primary end-point | 81 compared with muscle cover. 6,7 The study comparing biological matrix with synthetic mesh 8 concluded that there seems to be a higher incidence of implant loss when a biological matrix is used.
However, the synthetic mesh used wasnonabsorbable and titanized and therefore constitute an extra permanent foreign material. 8

| Study design and protocol
This study is a clinical, randomized, prospective trial described in the Gothenburg TIGR/Veritas Study protocol (ClinicalTrials. Gov identifier NCT02985073).

| Ethics and informed consent
The Regional Ethical Committee of Gothenburg reviewed and approved the study (189-16). Procedures followed were in accordance with the Helsinki Declaration of 1964, as revised, and the Good Clinical Practice guidelines. Personal data were treated in accordance with the General Data Protection Regulation.
Participants provided written informed consent to participate in the study.

| Recruitment of participants
The department is located in one of seven university hospitals in Sweden. All referrals for bilateral immediate breast reconstruction, to our department, were assessed for inclusion. All patients who met the inclusion criteria were asked for participation.

| Inclusion and exclusion criteria
Inclusion criteria were 18 years of age or older and indication for bilateral prophylactic mastectomy and immediate breast reconstruction. Exclusion criteria were inability to give informed consent, previous breast surgery, active smoking, and body mass index >30 kg/m 2 . Indication and surgical technique were discussed at a multidisciplinary team conference in all cases.

| Sample size
Assuming a target difference in overall complication frequencies between the groups of at least 35%, 32 breasts in each group would be needed to give the study 80% power, for a type I error rate of 5%.

| Randomization
The patients were operated with the biological mesh on one side and the synthetic mesh on the other side. During the operation, the patients were randomized, by the sealed envelope method, to which side the biological and the synthetic mesh, respectively, were going to be applied. The design was parallel and the intended allocation ratio in the groups was 1:1. The allocation sequence was concealed and a simple randomization approach was used. The patients were blinded to which mesh they received on which side.

| Interventions: Surgical technique and meshes/matrices
The surgical technique has been previously described 12,13 and was identical in the two breasts, with the exception of the mesh used. In ptotic breasts, a wise pattern skin resection was made, otherwise a submammary incision was performed. All mastectomies were nipplesparing. A sub-pectoral pocket was created and the inferior-medial and the inferior attachments of the major pectoral muscle were re- softer and more flexible. The slow-degrading part is completely resorbed after about three years. 17 Both meshes have been used in breast reconstruction previously. 12,15,[18][19][20][21][22] An anatomical tissue expander (TE) (CPX, Mentor Worldwide LLC) was used.
During the primary operation two suction drains (Exudrain, Mediplast), French gauge 14, were used for each breast, one subpectoral, and one subcutaneous. The drains were kept in place until the output was less than 30 ml per 24 h for 1 day, but for a maximum of 14 days.
Prophylactic perioperative and postoperative antibiotics (cloxacillin or clindamycin, in case of allergy) were given until the drains were removed. During the first two postoperative days a continuous infusion of ropivacaine 2 mg/ml, 4 ml/h, was given in each breast. After drain removal, clinically significant seromas were aspirated superficial to the TE injection port or with the aid of ultrasound. The patients were admitted for 2 days after the stage I operation. The patients wore a compression bra during the first weeks.
The first expander filling was performed after 2-3 weeks in our out-patient clinic. It was exchanged for a permanent implant (CPG, Mentor Worldwide LLC) about 3 months after the initial operation.
Stage II was performed as day-case surgery. 2.9 | Outcome: Complications during the first 12 months

| Clinical data collection
All patients were seen 1 and 2 weeks after stage I, during TE fillings, and before the exchange to the permanent implants. They were also seen 1 and 2 weeks and 3 and 12 months after stage II. Adverse events from the time of preparation for anesthesia until 12 months postoperation were registered. Systemic complications that could not be said to be related to one of the breasts were not included in the analysis. The definitions of complications are given in Table 2.

| Statistics
Descriptively, medians and ranges are given, except for daily volume which is given as a mean. Differences between samples from synthetic and biological meshes were analyzed using nonparametric  Thus, it became unethical to continue enrolling patients and the recruitment to the study was terminated before 32 breasts had been randomized in each group. No patient was lost to follow-up and 24 breasts in each group were available for 12 month-complication analysis ( Figure 1). As the synthetically and the biologically reconstructed breasts that were compared belonged to the same woman, systemic factors were exactly the same in the two groups (Table 3). Regarding breast-related factors (Table 4), there was one known case of invasive tumor in a breast subsequently randomized to a biological mesh, and one case of accidentally found an invasive tumor in breast randomized to synthetic mesh. All other mastectomies were prophylactic. All mastectomies were nipplesparing and the majority were performed via an inframammary incision ( Table 4). The most common complication was seroma formation with a frequency of 38% in the biological group and 3.8% in the synthetical group (p = .011) ( Table 5). All synthetic meshes were completely integrated during the exchange to a permanent implant, whereas the biological meshes were poorly integrated in the patients who had seroma (Figures 2-3). A higher frequency of total implant loss (stage I + II) could be seen in the biologic mesh group (8.5% vs. 2%), albeit not statistically significant (p = .083) ( Table 5). The two-implant losses, and re-operations, related to suspect penetration due to exceptionally thin mastectomy flaps were in the same patient.

| DISCUSSION
The present study investigates differences between a biological and a synthetic mesh regarding complication frequencies during the first 12 months after surgery. It is the first study to randomize between the two breasts in the same individual. The findings indicate that there is a higher risk for complications, particularly seroma and implant loss, when a biological mesh is used.
Risk factors for complications in immediate breast reconstruction include aspects related to the patient, to the surgical procedure, and to the oncological treatment. 23 A major strength of the present study is that the biological and synthetic mesh was used in the same patient. That is, patient-related factors, such as comorbidity, previous smoking, body mass index, breast size, and age, 23 Previous randomized controlled studies have indicated that the risk for implant loss is higher when a biological mesh is used, compared with traditional muscle cover 6 and compared with titanized nonabsorbable synthetic mesh 8 (Table 1). Our results are in accordance with these findings and indicate that the frequency of implant loss is higher in the biological mesh group ( | 85 three breasts (3/24, 12.5%) were affected in the biological mesh group and none in the synthetic mesh group. It has been hypothesized that the higher infection rates seen in biological meshes are secondary to seroma and that a higher postoperative seroma incidence is directly associated with the biological meshes. [29][30][31] Indeed, there was more seroma formation in the biological mesh group (Table 4). Given our results, it could also be hypothesized that the susceptibility to implant loss somewhat remains after stage II, as one implant was lost after stage II (Table 5). In summary, the results of the present study support that there is a higher incidence of seroma, secondary infection, and implant loss when biological meshes are used.
Although this a randomized controlled trial with identical background factors in the two groups, it has a few methodological weaknesses. Firstly, it was terminated as it became clear that the two meshes yield different esthetic results and therefore an asymmetry.
The esthetic results, and benefits and drawbacks with the two meshes in that respect, will be analyzed in a separate study. The plastic surgeons involved judged that it would be unethical to continue  (Table 5), albeit not statistically significant.
Secondly, some of the complications are still rare, although more common in the biological group. The sample size was calculated based on total complication frequencies and much larger samples would probably be required to establish differences in individual complications between groups, with high certainty. Thirdly, the study compares a single biological mesh with a single synthetical mesh.
Hence, the results might be biased by individual qualities of the meshes that are unrelated to whether they are biological or synthetical. Even so, as regards first-year complication frequencies, the safety of the meshes used in the present study has been demonstrated previously. 12,18 Therefore, there is no suspicion that any of the meshes used are of lower quality than other meshes currently on the market.
In conclusion, in the same patient, a synthetic mesh seems to yield a lower risk for serious complications, such as implant loss, than a biological mesh. Although very important, other factors such as costs and long-term esthetic and functional results, also need to be considered when the type of mesh is chosen. A longer follow-up of the patients will be performed in the future to address those factors.

ACKNOWLEDGMENTS
We would like to thank our former colleague Dr. Richard Lewin for participating in the initial planning of the study. The study was funded by grants from the federal government under the ALF agreement (ALFGBG-724171) and The Percy Falk Foundation for research into prostate cancer and breast cancer.