Sjogren Syndrome As A Risk Factor For Postoperative Complications Of Breast Reconstruction: A Propensity Score Matched Retrospective Cohort Study

PURPOSE: Sjögren syndrome (SS) is a chronic autoimmune disease estimated to affect between 1 and 4 million people in the US. Given the autoimmunity-mediated organ damages, chronic inflammation, and medication uses, we posited that SS is a risk factor for postoperative complications of breast reconstruction (BR). Despite the relative prevalence of SS, its potential significance as a risk factor for postoperative complications of breast reconstruction has, to our knowledge, not been studied at the population level. METHODS: The US Research Network of TriNetX, a large, multicenter database of real-time electronic health records, was leveraged to perform a propensity-score matched retrospective cohort study. Of patients who underwent BR, those with a prior history of SS comprised the exposure cohort; those without comprised the control cohort. Propensity score matching (1:1) was performed for demographic variables, medications, procedures, comorbidities, and body mass index classifications. Outcomes were postoperative rates of infection, seroma, bleeding, wound dehiscence, emergency department (ED) visits, and hospitalizations on postoperative day 90. Medications for SS were assessed by the Cox proportional hazards model. Exposures to these medications within 30 days prior to BR were used as covariates for outcomes on postoperative day 90. Bonferroni correction was used for the Cox regression, which lowered the significance level threshold to P 0.007. RESULTS: After propensity score matching, each cohort comprised 591 patients. On postoperative day 90, patients with SS were at significantly higher risks of postoperative infection (risk ratio [RR] 2.077, P-value 0.0243), wound dehiscence (RR 3.000, P-value 0.0013), and ED visits (RR 1.500; P-value 0.0276). According to the Cox regression analysis of patients with SS who underwent BR (N=618), preoperative exposure to cevimeline was significantly associated with wound dehiscence (hazard ratio 8.109, P-value 0.0006). CONCLUSION: Patients with SS undergoing BR may be significantly predisposed to postoperative infection, wound dehiscence, and ED visits. Prior to this study, the complication profile associated with SS had not been ascertained in the context of BR using a large, population-based analysis. Data presented herein may help to improve the quality of preoperative counseling and postoperative management for patients with SS. Moreover, the use of cevimeline in the preoperative window may be further investigated to optimize outcomes.