Quality of Life after Alloplastic versus Autologous Breast Reconstruction: The Influence of Patient Characteristics on Outcomes

After mastectomy, 13–69.5% of patients opt for breast reconstruction (BR) [1‒4]. BR has shown to play a significant role in restoring their physical and emotional well-being [5‒9]. Choosing between autologous (using the patient’s own tissue) and alloplastic (using implants) BR is a major decision for the patient. This choice depends on factors such as patient preferences, surgical suitability, health, body type, and cosmetic concerns [10].

Recent studies have consistently shown higher patient satisfaction and improved QoL following autologous BR [9‒19]. However, none of these studies adjusted for differences in baseline characteristics between groups. For example, higher BMI is required for autologous reconstruction for sufficient flap volume. And in cases of required adjuvant radiotherapy, in the Netherlands this is never administrated after autologous reconstruction, whereas in the alloplastic group, radiation can take place after BR. Consequently, many factors lead to indication bias in women seeking BR, which complicates evaluating patient satisfaction and QoL between various BR techniques.

This study investigates patient satisfaction and QoL over 3 years following BR, focusing on how baseline characteristics and treatment factors influence these outcomes. By adjusting for confounding factors, we aim to determine the true effect of the reconstruction technique on satisfaction and QoL.

The current study is a follow-up study of a previously published prospective longitudinal multi-centre cohort study, adding two additional follow-up timepoints [19]. Women over the age of 18 from three hospitals in the northern part of the Netherlands completed questionnaires at 5 timepoints: preoperatively and post-operatively at 6 weeks, 6 months, 1 year, and 3 years. Indication for mastectomy could be either therapeutic or prophylactic. A resume of the surgical course of every patient was made to investigate the total duration of the BR and to place the scores into the proper perspective, using the last intervention of the reconstruction as endpoint. Written informed consent was obtained from the patients before inclusion. The study was approved by the Medical Ethics Committee (2010.191).

Our primary outcomes were the following Breast-Q subscales: “Satisfaction with Breasts,” “Psychosocial,” “Sexual,” and “Physical Well-being” over time. The Dutch version of the Breast-Q reconstruction module is a validated, reliable, and reproducible questionnaire that assesses patient satisfaction and QoL after reconstructive surgery [20‒23]. Each Breast-Q scale is scored and converted into scores ranging from 0 to 100 and the minimal important difference is reported as 4 points [21, 22].

Online supplementary Table 1 (for all online suppl. material, see https://doi.org/10.1159/000543677) lists the patients, disease and treatment baseline characteristics, which were considered as potential confounders. The age of participants was recorded in years and the BMI was calculated as kg/m². Smoking status was noted as either “yes” or “no.” (Neo-)adjuvant therapies included radiotherapy, chemotherapy, and hormonal therapy. Comorbidities included hypertension, hypothyroidism, pulmonary disease, diabetes, and rheumatic disease. BR was either unilateral or bilateral and either performed immediately after mastectomy (immediate) or later (delayed). Complications were scored according to the Clavien-Dindo classification and were divided into 1–2 versus 3 or higher [24]. Two questionnaires were included for baseline measurement of overall well-being. The SF-36 covers eight domains of physical and emotional aspects giving a total score of physical and mental health with a minimally important difference of 5 points [25]. The HADS is specialized in detecting symptoms of anxiety and depression, categorized as normal (0–7) or abnormal (8–21) and has a minimally important difference of 1.5 points [26].

The sample size was calculated for the initial study [20]. As the research question was different, a new post hoc sample size calculation was performed using the linear multiple regression model in G*Power software [27]. The standard deviation of the “Satisfaction with Breasts” was used and the assumption that autologous BR scores 11 points higher than alloplastic BR 36 months after surgery. Post hoc power calculations gave a power of 84%, given an alpha of 0.05, a sample size of 35 per group, testing 12 factors, and an effect size of 0.32.

Using descriptive statistics, we described the demographic characteristics, and preoperative clinical data, stratified by reconstruction technique. The two groups were compared regarding baseline characteristics by using Student-t test, Mann-Whitney U test, or chi-square test, depending on type and distribution of the data. The mean Breast-Q scores per subscale were plotted stratified by reconstruction technique, showing the 5 assessments over time. The characteristics and scores of the dropouts were analysed to evaluate potential bias. When patients missed two or more assessments, they were defined as dropouts.

Linear mixed models were used to test the association between the reconstruction technique and Breast-Q scales. For each Breast-Q scale, two models were built to evaluate the differences of BR technique on QoL. “The original model” is the model used in most published studies; identification of confounders is based on literature and by assessing the AIC and log likelihood. This model is built for “Satisfaction with Breasts” and implemented to all Breast-Q subscales [20, 28]. The “adjusted model” was created for each Breast-Q scale separately. The confounders were identified by their influence on the effect size of reconstruction technique on QoL.

To create the adjusted model, our starting point included the most basic factors such as “autologous” versus “alloplastic BR,” “time,” and “preoperative baseline.” First, an additional variable was included: “BR technique x time” to evaluate if the difference in the Breast-Q subscales outcomes changes over time. Second, the analysis incorporated confounding factors making the effect size for BR technique on QoL change for ≥5% without compromising the model fit. All tested confounders are listed in online supplementary Table 1. Statistical analysis is performed using IBM SPSS Statistics for Windows (version 28.0) (IBM Corp., Armonk, NY).

In the 3 years follow-up, 37 out of the 76 patients (49%) who underwent alloplastic BR still participated in this study. For the autologous BR, this figure was 38 of the 63 (60%). See Figure 1 for a flowchart and overview of the patients participating at each timepoint per BR technique. Comparing patients who participated at all timepoints to dropouts showed no differences between the groups for patients’ characteristics as well as Satisfaction with Breasts score (online suppl. Tables 2, 3).

For autologous BR, 83% of the patients had additional operations such as asymmetry corrections, dogear corrections, and nipple reconstruction (either surgical and/or by tattooing). Within 2 years after the initial BR surgery, 89% of the women had completed their reconstruction, increasing to 95% 3 years after reconstruction.

Alloplastic reconstruction contained one-stage or two-stage reconstruction with secondary fine-tuning when needed (asymmetry or dogear corrections), change of the tissue expander for an implant in the two-stage group, followed by nipple reconstruction and/or tattooing when desired. In this group, 89% had more than one surgery, 90% of the patients had completed their BR after 2 years, and 97% after 3 years. Length of the surgical path was quite similar (14.3 vs. 13.7 months; online suppl. Table 4). Surgical complications grade III occurred significantly more in the autologous reconstruction group (online suppl. Table 5).

There were significant differences in patient and treatment characteristics between the participants in the alloplastic and autologous BR groups (Table 1). The autologous group was notably older (p = 0.029), had a higher average BMI (p < 0.001), less gene mutations (21% vs. 41%; p = 0.013), and more delayed BR (71% vs. 37%; p < 0.001). The timing of radiotherapy varied significantly between the groups (p < 0.001). Preoperative radiotherapy was more commonly given in the autologous BR group compared to the alloplastic BR group (43% vs. 9%). In contrast, no radiotherapy was administered after autologous BR, while 9% received post-surgery radiotherapy after alloplastic BR (p < 0.001). The time between chemotherapy and BR was longer in the autologous group (p = 0.009), whereas immediate pre-BR and post-BR chemotherapy was more common in the alloplastic group (p < 0.001). Furthermore, women obtained for autologous BR exhibited significant higher preoperative scores on the SF-36 (74 vs. 68) and experienced fewer symptoms of anxiety compared to those undergoing alloplastic BR (19 vs. 42) (p = 0.015, p = 0.002, respectively).

Patients in the autologous group scored higher on “Satisfaction with Breasts,” “Psychosocial Well-being,” “Physical Well-being,” and “Sexual Well-being” when adjusted for age, time post-surgery, baseline score, BMI, radiotherapy, immediate versus delayed BR, SF-36 scores, and anxiety (p < 0.001, p = 0.029, p = 0.090, p = 0.036, and p = 0.019, respectively) (Tables 2-5, right column).

Confounding factors identified in the analysis of “Satisfaction with Breasts” were depression, radiotherapy, indication for mastectomy, chemotherapy, and complications (Table 2, right column). Even after accounting for these factors, women undergoing autologous reconstruction consistently reported higher satisfaction with their breasts (B = 14.8 [6.1; 23.5], p < 0.001). The difference in satisfaction remained stable post-operatively (B = −0.1 [−2.8; 2.7], p = 0.966). Time after surgery was influencing the score for all women included (R = 3.2 [1.3; 5.0], p < 0.001).

Adjusting the model for the specific confounders influenced the differences between autologous and alloplastic BR patients, increased the correlation coefficient from 11.1 to 14.8 (Table 2, left and right column), indicating that the autologous BR group scores 14.8 points higher on “Satisfaction with Breasts” than the alloplastic group. “Psychosocial Well-being” was influenced by the administration of radiotherapy (B = −8.8 [−15.7; 1.8], p = 0.015), preoperative SF-36 scores (B = 0.4 [0.2; 0.6], p < 0.001), and time after surgery (B = 3.5 [1.6; 5.6], p < 0.001) (Table 3, right column). The adjusted model diminished the earlier reported difference between autologous and alloplastic BR to a non-significant level (Table 3).

Radiotherapy (R = −9.7 [−17.6; −1.9], p = 0.016) and the preoperative SF-36 (B = 0.2 [0.04; 0.4], p = 0.018) scores also played a role in “Sexual Well-being.” Timing and indication for surgery also played a role; women who underwent immediate BR had significantly lower sexual well-being compared to patients who underwent delayed BR (B = −17.0 [−25.4; −8.8], p < 0.001). Also, women who had prophylactic surgery scored higher on sexual well-being compared to those who had undergone a mastectomy for breast cancer (Table 4, right column). After adjusting the model for this specific subscale, no significant difference between autologous and alloplastic BR remained (Table 4).

In the subscale “Physical Well-being,” scores were significantly higher for women who underwent autologous BR (B = 10.1 [2.3; 17.8], p = 0.011), and this did not significantly change over time (B = −1.7 [−4.2; 0.7], p = 0.168) (Table 5, right column). Adjusting the model for this subscale had no influence on the significance of the difference between the two groups, although it increased the correlation (Table 5).

The plotted Breast-Q scores differences over time for each BR group remained stable and comparable in the post-operative course (Fig. 2). This was confirmed in the (adjusted) model. Integrating the variable describing difference between BR during follow up did not lead to a statistically significant difference. This means that the difference between both groups stayed the same between 6 weeks and 3 years post-operatively (Tables 2-5, right column).

To get more insight in the increased correlation from the original model to the adjusted model, the role of the identified confounders on “Satisfaction with Breasts” were plotted and analysed. This shows that high depression score before surgery significantly decreases the “Satisfaction with Breasts” score after autologous reconstruction, compared to alloplastic reconstruction at timepoint 6 weeks, 6 months and 3 years after reconstruction (p = 0.002, p = 0.011, p = 0.001, and p < 0.001, respectively) (Fig. 3a; Table 6). Without radiotherapy, the satisfaction remains significantly higher after autologous BR, but this difference seems to level after 3 years (p = 0.022, p = 0.019, p = 0.002, respectively; Fig. 3b; Table 6).

Former breast cancer patients score lower scores before the actual autologous BR but higher after 6 weeks, 1 year, and 3 years after autologous reconstruction (p = 0.004, p = 0.006, p = 0.017, and p = 0.020) (see Fig. 3c and Table 6). When no or mild complications (Clavien-Dindo <3) occur, autologous reconstruction patients score higher in satisfaction at all timepoints (p = 0.003, p < 0.001, p = 0.016, p < 0.001, p = 0.037) (Fig. 3d; Table 6).

This study examines Breast-Q changes over time following BR. The most important finding of this study is that the difference in “Satisfaction with Breasts” and “Physical Well-being” is not caused by the patient characteristics. It shows that, concerning “Satisfaction with Breasts,” autologous BR is superior to alloplastic BR directly after surgery until 3 years post-operatively. This difference has been found before but was thought to be related to factors such as a higher BMI, a lower prevalence of adjuvant therapy, and no two-stage reconstruction in the autologous group [10, 11, 29, 30]. Our new approach revealed that although BMI influences the overall “Satisfaction with Breasts” scores, it does not influence the difference in satisfaction between autologous and alloplastic BR. We identified baseline depression score, radiotherapy, indication for mastectomy, and Clavien-Dindo 3 or higher complications as important confounding factors [31]. These factors are partly correlated and represent characteristics, which deserve special attention. Preoperative depression is associated with complications and more common after breast cancer [32]. Also, radiotherapy will only be given after definitive breast cancer diagnosis. As all autologous reconstruction patients received RT before reconstruction, the lower score after surgery could be a delayed effect but could also be due to fatigue, or its influence on wound healing [33]. Further research is needed to investigate a possible correlation between these factors.

Another important finding is that the differences in “Psychosocial Well-being” and “Sexual Well-being” can be explained by patient and treatment characteristics. In both subscales, the preoperative baseline, the SF-36 score, and having had radiotherapy changed the correlation coefficient but also significantly influenced the Breast-Q score. In “Sexual Well-being,” women undergoing immediate BR scored preoperatively 17 points higher than women undergoing delayed BR, an important nuance often overlooked in studies lacking preoperative measurements or adopting different research questions and statistical analyses [9, 11, 12, 34]. The studies that included preoperative measurement had a different research question and therefore choose a different statistical analysis [19, 35]. In the majority of previous studies, autologous BR scored higher on “Satisfaction with Breasts” and “Psychosocial,” “Physical,” and “Sexual Well-being” than alloplastic BR. We are the first to correct for patient and treatment characteristics per scale based on their influence on the effect size of BR technique on QoL. Our finding is consistent with a published randomized BR study where women eligible for both techniques were randomized, excluding patients only eligible for one of the two techniques [36]. The findings of this study corroborate ours and also showed that only the difference in satisfaction with breast and physical well-being remains at the end of follow-up, although we reached this conclusion by using statistical analysis method, leaving more inclusive results for the general patient population [36].

None of the Breast-Q subscales showed a meaningful change in scores between the alloplastic and autologous groups between 6 weeks and 3 years, but over time, the overall scores for “Physical,” “Psychosocial Well-being” and “Satisfaction with Breasts” increased. A previous study showed that autologous BR patients still score higher on all Breast-Q scales 9–13 years after BR [29]. For the subscale “Satisfaction with Breasts,” autologous BR is definitely superior; however, our study showed that for the other subscales the outcomes are more nuanced. The earlier found differences over time could be explained by the effect of patient characteristics that are more prevalent in one of two groups, for example, fear for implant rupture, breast implant disease, capsular contracture, or long-term effects of oncological treatment [12, 37‒39]. We observed significantly lower Breast-Q scores preoperatively in most scales in the autologous group. In all Breast-Q scales, the superiority in score for alloplastic reconstruction shifts to autologous between baseline at 6 weeks after surgery. This underlines the difference in baseline characteristics. Alloplastic reconstruction patients mostly underwent immediate BR using tissue expanders and sometimes direct implants while autologous reconstruction surgery is mostly delayed. This explains the high increase in Breast-Q scores for the autologous BR group and the decrease for the alloplastic BR group at 6 weeks post-operatively. It also suggests that the intensity of autologous reconstructive surgery and the physical strain does not negatively influence QoL. Following this suggestion, we could look critically what risk of complications are affordable to choose for autologous reconstruction looking at long-term QoL.

Our study has several strengths. First, the main strength is its prospective nature including preoperative measurements; second, the addition of a 3-year follow-up is highly valuable as approximately 95% of patients have then completed their final reconstruction; third, the adjusted statistical models in which we corrected for patient characteristics; fourth, the use of the SF-36 to incorporate the general health of the patients as a confounding factor; fifth, the translation from the data to practical use [40]; and last, special attention to the inevitable risk of bias by dropouts.

There are also limitations. Due to the COVID-19 pandemic, elective surgeries, such as delayed BR and secondary corrections, were postponed, leading to long waiting lists. This may have resulted in a delayed completion of the surgical course [41, 42]. Analysing the dropouts showed that this had no effect on our results, although it decreased the power of our findings. This effected the possibilities we had in pinpointing the smaller differences. A larger cohort could also look at characteristics as type of mastectomy and degree of ptosis of the initial breasts. Lastly, an important difference between the two reconstruction techniques is the lack of a donor region in case of alloplastic reconstruction. Because the satisfaction with abdomen questionnaire was not applicable for the alloplastic group, this could not be included in the comparison.

Future prospective studies could also include no BR and other forms of BR such as the Goldilocks procedure and full BR using autologous fat transfer as these groups are growing and show different QoL patterns [42‒46]. As this study shows the superiority of QoL for autologous BR despite the complication risk, a study to investigate the criteria for this form of reconstruction could be in place.

Autologous BR stays superior to alloplastic BR when it comes to “Satisfaction with Breasts” and “Physical Well-being,” despite the higher risk of complications and longer recovery time. Patients who suited best for autologous reconstruction are women who needed prophylactic mastectomy, were without depression preoperatively, and had minimal risk of severe complications.

Our findings contribute valuable insights to the ongoing discussion surrounding BR techniques and their implications for patients’ well-being. In the context of shared decision-making, it is crucial for the physician to consider that baseline psychosocial health and sexual well-being, along with patient characteristics, play a significant role in quality of life. Opting for autologous reconstruction does not necessarily guarantee an improvement in all aspects of QoL as we have shown that, if corrected for baseline characteristics, “Psychosocial Well-being” and “Sexual Well-being” subscales were not superior.

We wish to thank all the patients and staff for their contribution to this study.

The study was approved by the Medical Ethics Committee University Medical Center Groningen (2010.191). Written informed consent was obtained from the patients before inclusion.

Paul M. N. Werker is a DMC member for Fidia Ltd, Milan-IT. Renumerations are used for research purposes. Iris L. Holt-Kedde, Nadia Sadok, Irene S. Krabbe-Timmerman, Geertruida H. de Bock, and Grigory Sidorenkov have nothing to disclose.

There are no funding sources to be declared.

I.L.H.-K., N.S., and I.S.K.-T. collected the data. G.B. and G.S. contributed to the analysis and interpretation of the results. I.L.H.-K. and N.S. took the lead in writing the manuscript. P.M.N.W. supervised the whole process. All authors provided critical feedback and helped shape the research, analysis, and manuscript.

The data that support the findings of this study are available from the corresponding author, I.L.H.-K., upon reasonable request.