Association of sociodemographic and oncological features with decision on implant-based versus autologous immediate postmastectomy breast reconstruction in Chinese patients

Abstract

Background and objectives: Immediate postmastectomy breast reconstruction (IPBR) has gained wide popularity in China. We sought to clarify the prevalence and predictors of implant-based vs autologous IPBR among Chinese patients.

Methods: A retrospective cohort study was performed using a prospectively maintained database. Women who underwent IPBR during 2001-2017 were included. The modality-specific trends were deciphered by curve fitting analysis. The association of sociodemographic and oncological features with the decision for implant-based vs autologous IPBR was investigated using multivariate logistic regression and structural equation modeling.

Results: Among 905 patients included in the study, 479 underwent implant-based IPBR and 426 underwent autologous procedures. The implant/autologous ratio has increased exponentially over time. Multivariate analysis demonstrated that unmarried patients with BMI ≤ 24 kg/m² , earlier clinical tumor stage, and preoperative pathological diagnosis of noninvasive lesion are more likely to choose implant-based IPBR compared to autologous procedures. The indirect effects of age, mastectomy type, and neoadjuvant chemotherapy were further demonstrated by the structural equations.

Conclusions: The sociodemographic and oncological features are directly or indirectly associated with the decision on type of IPBR. The findings may facilitate both patients and physicians to make a high-quality decision by holistic evaluation of the sociodemographic and oncological features.

Keywords: breast cancer; decision-making; immediate breast reconstruction; oncological features; sociodemographic characteristics.

Figure 1

Flow chart of patient inclusion. The decay in the cohort is shown from initial enrollment of patients through the selection of the cases for the ultimate analysis in this study. Abbreviation: BR, breast reconstruction

Figure 2

Temporal trends in the use of implant‐ vs autologous‐based immediate postmastectomy breast reconstruction. The implant/autologous ratio increased in an exponential manner (R ² = 0.87, P < 0.01) among the patients in the current cohort

Figure 3

Multivariate logistic regression analysis. (A) The likelihood of having autologous‐(1) vs implant‐based (0) breast reconstruction was compared (n = 905). BMI, marital status, clinical tumor staging, and preoperative pathological diagnosis were independently associated with the decision on type of breast reconstruction in the adjusted model. (B) The sensitivity and specificity of the regression model were determined by receiver operating characteristic (ROC) curve. The areas under curve (AUC, 0.766; 95% CI, 0.725‐0.806) suggested satisfying model fitting

Figure 4

Structural equation models. (A) Hypothesized model. The hypotheses were rejected due to poor model fit. (B) Final model. The model with good fit coefficients shows that the sociodemographic characteristics can represent age, BMI, and marital status; while the oncological features can serve as a proxy for clinical tumor staging, preoperative pathological diagnosis, type of mastectomy, and neoadjuvant chemotherapy. The decision on type of immediate postmastectomy breast reconstruction (IPBR) is significantly affected by the sociodemographic characteristics and oncological features and the impact of the sociodemographic characteristics on the decision‐making is partly mediated by the oncological features. Note: β values refer to standardized direct effects on the downstream variables, and a values refer to the standardized factor load capacity. The solid lines with single arrow represent significant parameter estimates, and the dotted lines represent nonsignificant parameter estimates. Error variances and covariances are not shown. ^#factor load was defined as 1.00 in the unstandardized estimates. *P < 0.05, **P < 0.01