Does Higher Intraoperative Fraction of Inspired Oxygen Improve Complication Rates Following Implant-Based Breast Reconstruction?

Abstract Background The surgical literature debates about whether an average intraoperative fractional inspired level of oxygen (FiO2) greater than 80% confers lower postsurgical complication rates. Although some evidence demonstrates minimal or no difference in short-term mortality or surgical site infections, few studies suggest negative long-term outcomes. Objectives To the best of our knowledge, this is the first study examining the relationship between intraoperative FiO2 levels and postoperative outcomes in the setting of immediate prepectoral implant-based breast reconstruction. Methods The authors retrospectively reviewed the complication profiles of 309 patients who underwent prepectoral 2-stage breast reconstruction following mastectomy between 2018 and 2021 at a single institution. Two cohorts were created based on whether intraoperative FiO2 was greater than 80% or less than or equal to 80%. Complication rates between the cohorts were analyzed using Chi-squared test, Fisher’s exact test, and multivariable logistic regressions. Variables examined included demographic information; smoking history; preexisting comorbidities; history of chemotherapy, radiation, or axillary lymph node dissection; and perioperative information. Results Chi-squared and multivariable regression analysis demonstrated no significant difference between cohorts in complication rates other than reoperation. Reoperation rates were significantly increased in the FiO2 greater than 80% cohort (P = 0.018). Multivariable logistic regression also demonstrated that the use of acellular dermal matrix was significantly associated with increased postoperative complications (odds ratio 11.985; P = 0.034). Conclusions Complication rates did not statistically differ in patients with varying intraoperative FiO2 levels outside of reoperation rates. In the setting of implant-based prepectoral breast reconstruction, hyperoxygenation likely does not lead to improved postsurgical outcomes. Level of Evidence: 3

The fractional level of inspired oxygen (FiO 2 ) is the concentration of oxygen in the gas mixture that is inhaled. 1 Current literature in general surgery and anesthesiology hypothesizes that intraoperative FiO 2 greater than 80% during endotracheal intubation confers lower postsurgical complication rates. 2,3 In theory, increased FiO 2 augments oxygenation during perfusion to benefit wound healing and nutrient delivery to tissues as part of the stress response in the operative setting. While the debate currently centers around whether an FiO 2 of greater than 80% will offer no benefit, short-term benefit or, conversely, long-term complications due to hyperoxygenation, these studies have only been evaluated in the settings of abdominally based surgeries and non-thoracic surgeries. [2][3][4] In the emergency setting, higher FiO 2 contributed to lower surgical site infection (SSI) rates in colorectal surgery. 3 Other studies that broadened the scope of outcomes concluded that increased FiO 2 decreased SSIs in any abdominal-based surgery in the emergency setting but did not confer clinical advantages in the non-emergent setting. 2 The nidus for our study, thus, is whether FiO 2 offers clinical advantage in the setting of postoperative outcomes following breast reconstruction.
In broad strokes, the literature demonstrates conflicting perspectives as studies both challenge and support the benefit of increased intraoperative oxygenation. In the context of non-emergent abdominal surgery, increased FiO 2 did not result in lower SSIs or adverse outcomes including myocardial ischemia and atelectasis when compared with conventional FiO 2 (near 30%). 5 Other studies have similarly shown no significant short-term mortality advantage in the setting of non-thoracic surgeries. 6 Recent surgical and critical literature studies demonstrate that increased FiO 2 may both reduce and increase the risk of atelectasis, highlighting the starkly oppositional viewpoints that exist from a pulmonary perspective. 3,4,7,8 Meta-analyses and systematic reviews additionally offer conflicting viewpoints. In a meta-analysis of 28 randomized control trials examining the role of increased vs standard intraoperative FiO 2 on outcomes related to mortality, SSI, respiratory insufficiency, serious adverse events, and length of stay, there was no evidence to suggest a reduction in SSIs, but adverse events including mortality may increase at higher levels of intraoperative FiO 2 . 9 An additional systematic review of 23 studies concluded that increased FiO 2 did not decrease SSIs but may increase long-term adverse respiratory and cardiac outcomes. 1 Our study focuses on whether intraoperative FiO 2 is correlated with short-term postoperative complications within the first 6 months following stage 1 of reconstruction, specifically in the setting of breast reconstruction. To the best of our knowledge, this is the first study on the role of intraoperative FiO 2 in plastic and reconstructive surgery, which may offer insight into the utility of potentially cost-effective intraoperative measures.

Study Design
This study retrospectively reviewed the outcomes of female patients who underwent staged prepectoral breast reconstruction between January 2017 and October 2021 at a single institution. All reconstructions were performed by author B.T. and 3 other surgeons at this tertiary care center. The study was approved by the Stanford University IRB (IRB #49366), and charts were accessed by author E.T. All data collection was conducted in a de-identified manner. Patients were divided into 2 cohorts based on an average intraoperative FiO 2 of greater than or equal to 80% or less than 80% using endotracheal intubation.
Demographic, comorbidity, and perioperative information collected included age, BMI, smoking history, diagnosis of diabetes mellitus or hypertension, American Society of Anesthesiologists Classification score (ASA), use of acellular dermal matrix (ADM), and mastectomy incision patterns. Age was categorized into 3 groups: less than 40 years of age, between 40 and 50 years of age, and above 50 years of age. Age categorizations were chosen in with respect to national averages of women undergoing mastectomy and reconstruction. BMI was categorized into underweight, normal, overweight, and obese groups. ASA score was categorized according to whether or not patients had an underlying severe systemic disease that warranted a physical status indicator of 3 or greater. Examples of systemic diseases that qualify for a score of 3 include severe heart disease and diabetes with vascular complications.
Charts were also reviewed for incidence of postoperative complications up to 6 months following stage 1 of reconstruction. Complications included infection, seroma, Therapeutic hematoma, ischemia and necrosis of the mastectomy skin flaps, nipple necrosis, explantation of the tissue expander, dehiscence, and reoperation.

Statistical Analysis
Data were recorded in a de-identified manner and organized into tables using Microsoft Excel (Microsoft Inc., Seattle, WA). Chi-squared and Fisher's exact tests were used to assess the variation in demographic characteristics and complication profiles between the 2 cohorts. Upon statistically significant variation between the cohorts in reoperation rates, multivariable logistic regressions were performed to calculate serially adjusted odds ratios for undergoing reoperation following the index procedure. A second multivariable logistic regression was used to evaluate predictors of experiencing at least 1 postoperative complication. Covariates included FiO 2 greater than 80%, age, BMI, smoking history, diagnosis of diabetes mellitus or hypertension, ASA, use of ADM, and mastectomy incision patterns. P-values < 0.05 were considered statistically significant. All analyses were completed using Stata, version 16.1 (StataCorp, LLC, College Station, TX).

RESULTS
Three hundred-nine female patients underwent staged prepectoral breast reconstruction between January 2017 and October 2021. Patients were divided into 2 cohorts: patients who had an intraoperative FiO 2 of less than or equal to 80% (n = 114) and patients with breast reconstruction who had an FiO 2 of greater than 80% (n = 195). Table 1 presents no significant difference in age, BMI, smoking history, diagnosis of diabetes or hypertension, ASA classification, use of ADM, and mastectomy incision patterns between the FiO 2 greater than 80% cohort and FiO 2 less than or equal to 80% using multivariate regression. Table 2 compares complication rates between the 2 cohorts. There was no significant difference between cohorts in rates of postoperative infection, seroma, hematoma, mastectomy flap ischemia and necrosis, nipple necrosis, explantation of implants, and wound dehiscence. However, rates of reoperation were significantly higher in the FiO 2 greater than 80% group compared with the FiO 2 less than 80% group (16.4% vs 7.0%; P = 0.018). In a multivariable logistic regression evaluating predictors of undergoing reoperation, including FiO 2 greater than 80%, age, increased BMI, smoking history, diagnosis of diabetes mellitus or hypertension, ASA score, use of ADM, and mastectomy incision patterns, patients in the FiO 2 greater than 80% cohort had significantly higher odds of reoperation (odds ratio [OR] 2.367; P = 0.043) ( Table 3). The use of ADM was omitted from this model as no patients without ADM underwent reoperation in this dataset. Table 4 presents a multivariable logistic regression evaluating predictors of experiencing one or more postoperative complications that included the following covariates: FiO 2 greater than 80%, age, BMI, smoking history, diagnosis of diabetes mellitus or hypertension, ASA score, use of ADM, and mastectomy incision patterns. Only the use of ADM revealed a significant association with increased postoperative complications following stage 1 reconstruction (OR 11.985; P = 0.034).

DISCUSSION
SSIs and other postoperative complications are a significant cost burden in healthcare systems. Increased intraoperative FiO 2 (above 80%) has been proposed as a cost-effective measure to potentially decrease postoperative complications but has yet to be studied within the context of breast reconstruction. The currently available general surgery and anesthesia literature is engaged in an active debate over whether increased intraoperative FiO 2 contributes to reduced postoperative complication rates. Tissue oxygenation has been implicated in the stimulation of fibroblasts, collagen deposition, and epithelialization. 10 It has been hypothesized that, because wound beds contain significantly less oxygen postoperatively, reduced oxygenation may contribute to impaired ability to utilize oxidative bursts to decrease bacterial load and facilitate wound healing. 9 Hypermetabolic tissue within the wound bed has increased oxygen demand relative to supply, creating a hypoxic environment that leads to increased oxidative stress. 11 Studies of ongoing hyperoxygenation demonstrate that increased reactive oxygen species negate the antioxidant properties of well-oxygenated tissue and result in damage at both the cellular and organ levels. 12 By contrast, however, hypoxia notably induces transcription factors including hypoxia-inducible factor 1-alpha (HIF1-α), which promote angiogenesis and cell proliferation in hypoxic tissues. 13 Given the varied outcomes of tissue oxygenation at the molecular level, it is conceivable that postoperative outcomes are also varied.
Increased intraoperative FiO 2 in the setting of nonreconstructive surgical procedures has demonstrated mixed results. FiO 2 greater than 80% only consistently contributed to lower SSIs in cases of emergent abdominalbased (eg, colorectal) surgeries. 2,3 In the non-emergent abdominal-based surgery setting, increased FiO 2 did not result in reduced SSIs or in increased adverse outcomes. 5 Additional studies echo no significant clinical advantage concerning short-term mortality in non-thoracic surgeries. 4 Conclusions regarding respiratory insufficiency secondary to atelectasis are also variable. Studies have demonstrated that increased FiO 2 both precipitates and prevents atelectasis. 7,8 The findings in our study demonstrate that, in the setting of prepectoral breast reconstruction, increased intraoperative FiO 2 did not decrease postoperative rates of infection, seroma, hematoma, mastectomy flap ischemia and necrosis, nipple necrosis, implant explantation, and dehiscence, when compared with a similar cohort of patients with breast reconstruction who had an intraoperative FiO 2 of less than or equal to 80%. Our data did, however, demonstrate increased rates of reoperation in the cohort of patients who had an intraoperative FiO 2 of greater than 80%. While some of the aforementioned literature suggests that oxidative damage due to an average intraoperative hyperoxygenation of greater than 80% can increase the risk of postoperative complications, reoperation is unlikely to be clinically significant without additional co-existing complications such as increased rates of infection or dehiscence. This is likely an artifact in our dataset. However, a multivariate regression using FiO 2 along with additional predictors of one or more complications demonstrated that only the use of ADM was a significant predictor of complications. There is a plethora of literature that argues for and against ADM in the setting of increased postoperative complications, particularly infection and seromas, and warrants discussion outside the framework of FiO 2 . [14][15][16][17][18][19] While it has been hypothesized that manipulation of intraoperative FiO 2 can be a cost-effective measure to prevent or reduce postoperative complications, our data demonstrate no evidence of clinical benefit in the setting of breast reconstruction. Our study was limited, however, to prepectoral breast reconstruction and was limited to a single-institution database that did not control for intrainstitutional surgeon-specific complications. A larger sample size and further gradation of intraoperative FiO 2 into categories beyond a binary division of greater than and less than or equal to 80% FiO 2 may reveal more nuanced effects of FiO 2 on postoperative outcomes. Additionally, as other studies have aptly demonstrated, separating intraoperative FiO 2 from the effects of postoperative antibiotics can be challenging but would likely reveal additional nuanced findings if studied.

CONCLUSIONS
Literature in the surgical and anesthesia spheres continue to debate the utility and safety of manipulating intraoperative FiO 2 to levels above the standard (30%). To our knowledge, complication rates following reconstructive surgery have not been studied in the context of varying intraoperative FiO 2 levels. Our data demonstrated that there was no significant difference in complication rates following prepectoral breast reconstruction in patients who had an intraoperative FiO 2 of greater than 80% vs less than or equal to 80% except for rates of reoperation. It is unlikely that increased rates of reoperation in the FiO 2 greater than 80% cohort are clinically relevant without other statistically significant complication differences between cohorts such as infection rates and rates of dehiscence. At this time, we believe that hyperoxygenation likely does not lead to improved postsurgical outcomes in the setting of breast reconstruction. Further study is warranted with larger sample sizes, various reconstruction procedures, and a larger gradation of intraoperative FiO 2 levels.

Disclosures
The authors declared no potential conflicts of interest with respect to the research, authorship, and publication of this article. One hundred ninety-one patients experienced at least one complication due to the index procedure. CI, confidence interval; FiO 2 , fractional level of inspired oxygen; NSM, nipple-sparing mastectomy; OR, odds ratio.