Impact of glycaemic control on complication rates after breast reconstruction

Abstract

Full text

While breast reconstruction (BR) remains a cornerstone of breast cancer care by restoring physical form and psychological well-being, safely extending these benefits to high-risk patients with increasingly complex comorbidities, particularly diabetes, remains a significant challenge^1^.

In this retrospective cohort study, the National Surgical Quality Improvement Program (NSQIP) (2021–2023) was queried to examine whether preoperative glycaemic control, quantified by haemoglobin A1c (HbA1c), was associated with 30-day postoperative morbidity following single-stage BR. NSQIP is a validated and risk-adjusted surgical quality improvement registry that aggregates outcomes data from hundreds of participating hospitals worldwide. The authors identified female breast cancer patients with diabetes and documented HbA1c undergoing single-stage autologous (ABR) or implant-based breast reconstruction (IBR) via International Classification of Diseases (ICD)/Current Procedural Terminology (CPT) codes^1–2^. Multistage procedures and incomplete records were excluded. Outcomes were assessed over the standardized 30-day postoperative period. Continuous and categorical variables were compared using the Student’s t test and χ^2^ analysis, respectively. Statistical significance was defined as a two-tailed P < 0.05.

Of the 739 patients meeting inclusion criteria, 606 (82%) underwent IBR and 133 (18%) underwent ABR. Patients were stratified according to the American Diabetes Association’s^3^ threshold for poor glycaemic control (HbA1c ≥ 6.5%). Among patients undergoing IBR, 55% exhibited HbA1c ≥ 6.5%. Baseline comorbidities, including ASA class, obesity, and hypertension, were comparable, apart from higher smoking prevalence in the high HbA1c group (P = 0.014). Notably, the composite 30-day complication rate—encompassing mortality, reoperation, readmission, and both surgical and medical complications—did not differ significantly between glycaemic control groups in the IBR cohort (15.0% versus 15.4%; P = 0.90). Table 1 provides an overview of all extracted variables.

In contrast, outcomes diverged markedly in the ABR cohort. Of the 57 patients with HbA1c ≥ 6.5%, 43.9% experienced any complication, compared with 25.0% among those with HbA1c < 6.5% (P = 0.022). Flap selection also varied. Free flaps were more common in the control group (86.8% versus 68.4%; P = 0.010) and latissimus dorsi flaps more common in the high HbA1c group (21.1% versus 7.9%; P = 0.028).

This procedure-specific susceptibility of ABR to dysglycaemia may reflect the greater physiological burden of autologous procedures. ABR demonstrated significantly longer operative times than IBR (mean(standard deviation) 425(176) versus 205(85) minutes; P < 0.001)), which, alongside greater tissue trauma, amplifies metabolic stress. Under these conditions, poorly controlled diabetes may further compromise host defence, delay wound healing, and predispose to postoperative complications^2^. Conversely, the shorter duration and lower invasiveness of IBR may attenuate the impact of suboptimal glycaemic control on early outcomes^4^.

Nearly half of patients with HbA1c ≥ 6.5% who underwent ABR developed a complication within the first postoperative month, raising concerns about proceeding with flap procedures in this subgroup without adequate metabolic optimization. Although these data should be interpreted with appropriate caution, they highlight the potential importance of procedure-specific risk stratification rather than a ‘one-size-fits-all’ approach. IBR might represent a comparatively safer option when surgical timelines are urgent or glycaemic control remains suboptimal; however, further research is essential.

This study is limited by the 30-day NSQIP follow-up window, which precludes assessment of delayed complications like capsular contracture or late implant loss. Although the authors’ analysis did not identify an association between elevated HbA1c and early morbidity following single-stage implant-based reconstruction, Capito et al. ^5^ reported increased implant loss among patients with diabetes undergoing two-stage reconstruction, suggesting that glycaemic dysregulation may exert delayed effects not captured within the NSQIP window. In addition, the registry does not allow for complication grading. Furthermore, granular intraoperative or reconstructive details, such as implant positioning or flap perfusion parameters, were unavailable, limiting mechanistic insight. Future studies are warranted to further elucidate the relationship between glycaemic control and postoperative outcomes following breast reconstruction.

As breast cancer survivorship rises and type 2 diabetes becomes increasingly prevalent, reconstructive surgeons will be challenged to navigate complex risk profiles with limited evidence-based guidance. The present study represents a step toward individualized, data-driven surgical decision-making by identifying potential procedure-specific metabolic vulnerabilities.