Advertisement

Impact of opioid-free anesthesia on complications after deep inferior epigastric perforator flap surgery: A retrospective cohort study

  • Harold Mulier
    Affiliations
    KU Leuven – University of Leuven, Herestraat 49, B-3000 Leuven, Belgium
    Search for articles by this author
  • Bob De Frene
    Affiliations
    Department of Plastic and Reconstructive Surgery, AZSint Jan AV Brugge-Oostende, Campus Brugge, Ruddershove 10, B-8000 Brugge, Belgium
    Search for articles by this author
  • Lara Benmeridja
    Affiliations
    Department of Plastic and Reconstructive Surgery, AZSint Jan AV Brugge-Oostende, Campus Brugge, Ruddershove 10, B-8000 Brugge, Belgium

    Department of Plastic and Reconstructive Surgery, Ghent University Hospital, B-9000 Ghent, Belgium
    Search for articles by this author
  • Florian Vanhoorebeeck
    Affiliations
    Department of Anesthesiology, Intensive Care and Reanimation, AZSint Jan Brugge-Oostende, Campus Brugge, Ruddershove 10, B-8000 Brugge, Belgium
    Search for articles by this author
  • Bruno Denis
    Affiliations
    Department of Anesthesiology, Intensive Care and Reanimation, AZSint Jan Brugge-Oostende, Campus Brugge, Ruddershove 10, B-8000 Brugge, Belgium

    Department of Anesthesiology and Acute Pain Service, Cliniques Universitaires Saint Luc, University Catholic of Louvain, av Hippocrate 10, B-1200 Brussels, Belgium
    Search for articles by this author
  • Bob Casaer
    Affiliations
    Department of Plastic and Reconstructive Surgery, AZSint Jan AV Brugge-Oostende, Campus Brugge, Ruddershove 10, B-8000 Brugge, Belgium
    Search for articles by this author
  • Fabrice Jacques Rogge
    Affiliations
    Department of Plastic and Reconstructive Surgery, AZSint Jan AV Brugge-Oostende, Campus Brugge, Ruddershove 10, B-8000 Brugge, Belgium
    Search for articles by this author
  • Kris Leleu
    Affiliations
    Department of Anesthesiology, Intensive Care and Reanimation, AZSint Jan Brugge-Oostende, Campus Brugge, Ruddershove 10, B-8000 Brugge, Belgium
    Search for articles by this author
  • Jan Mulier
    Correspondence
    Corresponding author at: Department of Anesthesiology, Intensive Care and Reanimation, AZSint Jan Brugge-Oostende, Campus Brugge, Ruddershove 10, B-8000 Brugge, Belgium.
    Affiliations
    KU Leuven – University of Leuven, Herestraat 49, B-3000 Leuven, Belgium

    Department of Anesthesiology, Intensive Care and Reanimation, AZSint Jan Brugge-Oostende, Campus Brugge, Ruddershove 10, B-8000 Brugge, Belgium

    UGhent – University of Ghent, B-9000 Ghent, Belgium
    Search for articles by this author
Open AccessPublished:September 30, 2020DOI:https://doi.org/10.1016/j.bjps.2020.09.004

      Summary

      This study measured the number of complications after deep inferior epigastric perforator (DIEP) flap reconstruction performed under opioid-free anesthesia (OFA) combined with goal-directed fluid therapy or opioid anesthesia with liberal fluid therapy (OA). This retrospective cohort study consisted of 204 patients who underwent DIEP flap reconstruction at AZSint Jan Brugge between April 2014 and March 2019. Primary outcomes were complications, according to the Clavien–Dindo classification and the length of hospital stay (LOS). The secondary outcomes were flap failure, postoperative nausea and vomiting (PONV), postoperative pain, postoperative opioid consumption, and postoperative skin flap temperature. OFA included a combination of dexmedetomidine, lidocaine, and ketamine without any opioid administered pre- or intraoperatively. OA included a combination of sufentanil and remifentanil. OFA patients received strict goal-directed fluid therapy, whereas OA patients received liberal fluids to maintain perfusion pressure. All patients except 7 (TIVA with remifentanil) received inhalation anesthesia combined with an infusion of propofol. Of the 204 patients, 55 received OFA and 149 received OA. There were no differences in major complications, but fewer minor complications in the OFA group (17.9% vs. 51.4% and P < 0.001). Flap failure occurred in three patients of the OA group. Six patients developed flap thrombosis (five OA patients and one OFA patient). OFA was associated with fewer postoperative opioids, shorter LOS, less PONV, and less pain. In patients without previous nausea, the PONV incidence was higher in the OA group than in the OFA group (12.7% vs. 43.6% and P < 0.001). Patients with previous nausea more frequently required postoperative opioids and had a nausea rate of 60.87%.

      Keywords

      Introduction

      The reconstruction of deep inferior epigastric perforator (DIEP) flaps
      • Wade R.G.
      • Razzano S.
      • Sassoon E.M.
      • et al.
      Complications in DIEP flap breast reconstruction after mastectomy for breast cancer: a prospective cohort study comparing unilateral versus bilateral reconstructions.
      ,
      • Siegel R.L.
      • Miller K.D.
      • Jemal A
      Cancer statistics.
      involves the transfer of free abdominal tissue to the breast using microsurgery.
      • Blondeel P.N
      One hundred free DIEP flap breast reconstructions: a personal experience.
      ,
      • Allen J.
      The evolution of perforator flap breast reconstruction: twenty years after the first DIEP flap.
      Reducing perioperative opioids may enhance recovery after surgery (ERAS), as measured by the number of minor and major complications and the length of hospital stay (LOS), and is therefore one of the principles of ERAS.
      • Batdorf N.J.
      • Lemaine V.
      • Lovely J.K.
      • et al.
      Enhanced recovery after surgery in microvascular breast reconstruction.
      Opioid-free anesthesia (OFA) reduces postoperative nausea and vomiting (PONV),
      • Mulier J.
      • Wouters R.
      • Dillemans B.
      • De Kock M
      A randomized controlled, double-blind trial evaluating the effect of opioid-free versus opioid general anaesthesia on post-operative pain and discomfort measured by the QoR-40.
      the most common minor complication affecting flap perfusion.
      • Manahan M.A.
      • Basdag B.
      • Kalmar C.L.
      • et al.
      Risk of severe and refractory postoperative nausea and vomiting in patients undergoing DIEP flap breast reconstruction.
      Major complications such as flap failure are rare but can have devastating effects.
      • Batdorf N.J.
      • Lemaine V.
      • Lovely J.K.
      • et al.
      Enhanced recovery after surgery in microvascular breast reconstruction.
      Many perioperative factors, including low cardiac output, hypothermia, and surgical stress, may cause vasoconstriction or thrombosis, thereby reducing flap blood flow.
      • Hagau N.
      • Longrois D.
      Anesthesia for free vascularized tissue transfer.
      OFA may more effectively lower perioperative stress but can induce more hypotension, making the impact on flap perfusion uncertain. Dexmedetomidine

      Li, B.,Li Y., Tian S.et al. Anti-inflammatory effects of perioperative dexmedetomidineadministered as an adjunct to general anesthesia: a meta-analysis. Sci. Rep. 5, 12342.

      ,
      • Byon H.
      • Ok S.
      • Lee S.H.
      • et al.
      Dexmedetomidine inhibits phenylephrine-induced contractions via alpha-1 adrenoceptor blockade and nitric oxide release in isolated rat aortae.
      and lidocaine,
      • Cassuto J.
      • Sinclair R.
      • Bonderovic M.
      Anti-inflammatory properties of local anesthetics and their present and potential clinical implications.
      ,
      • Gherardini G.
      • Gurlek A.
      • Cromeens D.
      • et al.
      Drug-induced vasodilation: in vitro and in vivo study on the effects of lidocaine and papaverine on rabbit carotid artery.
      which are common components of OFA, have anti-inflammatory and vasodilatory effects that may improve free-flap perfusion.
      This study tested the hypothesis that OFA combined with goal-directed fluid therapy for DIEP flap surgery is associated with fewer complications.

      Patients and methods

      In this cohort study, we included all patients in our hospital database who underwent DIEP flap surgery between January 2014 and April 2019. All patients entering our hospital provide their consent to allow retrospective data analysis. The study protocol was approved by our institutional review board and registered at http://clinicaltrials.gov. This manuscript adheres to the STROBE and CONSORT guidelines.
      At our institution, the possibility of breast reconstruction is discussed with all women undergoing mastectomy for established breast cancer or prophylactically as a risk reduction strategy, and the decision regarding the timing and type of reconstruction is based on each patient's preference and surgical options. Approximately 80% of patients requesting breast reconstruction choose a DIEP flap. When adjuvant radiotherapy is required, our breast multidisciplinary team policy is to offer delayed autologous tissue reconstruction. Patients with a high body mass index (BMI; >30 kg/m2) are advised to lose weight, and those who smoke are advised to stop smoking at least 4 weeks before surgery; however, neither of these characteristics will exclude women from being offered DIEP flap surgery. Hormone therapy, when a component of the breast cancer treatment is stopped one week before surgery and continued after discharge. Prior to DIEP flap surgery, all women undergo computed tomography angiography with three-dimensional image reconstruction for an assessment of the lower abdominal wall perforators and recipient vessels by an experienced radiologist.

       Methods

      All DIEP flap reconstructions were performed by senior surgeons using a standard technique.
      • Allen J.
      The evolution of perforator flap breast reconstruction: twenty years after the first DIEP flap.
      These surgeons routinely preserve the superior inferior epigastric vein for “supercharging” if venous congestion occurs.
      • Rothenberger J.
      • Amr A.
      • Schiefer J.
      • Schaller H.
      • Rahmanian-Schwarz A
      A quantitative analysis of the venous outflow of the deep inferior epigastric flap (DIEP) based on the perforator veins and the efficiency of superficial inferior epigastric vein (SIEV) supercharging.
      The internal thoracic vessels at the level of the third or fourth rib were the first-choice recipient vessels. End-to-end anastomoses were performed using 9-0 or 10-0 Nylon (ST&T), depending on the vessel caliber.
      The choice of OFA or OA was at the discretion of the attending anesthesiologist, only some of whom were experienced with OFA. OFA was defined as the administration of no opioids either pre- or intraoperatively until wound closure, and opioids were only given when the patient was fully awake and had been administered nonopioid analgesics previously. OA was defined as the administration of any opioid during the pre- or intraoperative period. Postoperative opioids (after OFA or OA) were defined as the administration of opioids from the time of wound closure until discharge from the hospital.
      Most patients received a balanced general anesthetic, which was usually a combination of 0.5 MAC sevoflurane inhalation and a continuous propofol infusion at 3 mg/kg/h. Seven patients in the opioid group received total intravenous anesthesia (TIVA with remifentanil) without inhalation. Neuromuscular blockade (NMB) was achieved by the administration of rocuronium or cisatracurium in a continuous infusion to achieve periods of deep NMB when required.
      The method by which OFA was administered consisted of combining sympatholytics based on monitoring the analgesia nociception index (ANI).
      • Gazi M.
      • Abitagaoglu S.
      • Turan G.
      • et al.
      Evaluation of the effects of dexmedetomidine and remifentanil on pain with the analgesia nociception index in the perioperative period in hysteroscopies under general anesthesia. A randomized prospective study.
      Dexmedetomidine was administered as follows: 0.3 mcg/kg 15 min before induction, 0.1 mcg/kg at induction, and an infusion of 0.1 mcg/kg/h for maintenance. Lidocaine was administered as 1 mg/kg at induction, followed by an infusion of 1 mg/kg/h for maintenance. Ketamine was administered as 0.1 mg/kg at induction, 0.7 mg/kg (maximum 50 mg) before incision, and an infusion of 0.1 mg/kg/h for maintenance. Postoperative analgesia was further improved by continuing very low doses of dexmedetomidine (0.05 mcg/kg/h), ketamine (0.05 mg/kg ideal body weight/h), and lidocaine (0.5 mg/kg/h) for the first few hours (maximum, 5 h) after surgery, with maximal bolus doses of 10 mg lidocaine, 1 mg ketamine, and 1 mcg dexmedetomidine administered every 15 min on an as-needed basis.
      Sufentanil was administered as the OA at a dose of 0.1–0.3 mcg/kg at induction, followed by additional 0.1–0.2-mcg/kg doses as needed during surgery or by a continuous infusion of remifentanil at a dose of 0.20–0.35 mcg/kg/h. Since 2014, various adjunct medications (e.g., clonidine, dexmedetomidine, ketamine, and lidocaine) have been administered as a single dose to reduce the total dose of intraoperative opioids, and this still classifies as receiving OA. Both groups received paracetamol and ketorolac postoperatively.
      All patients undergoing OFA were managed with goal-directed fluid therapy, receiving fewer than 1200 mL of intravenous (iv) fluids intraoperatively. Patients undergoing OA received more liberal fluid therapy. The OFA group consists therefore of a combination of zero intraoperative opioids with goal-directed fluid therapy.

       Data collection

      The following demographic data were retrieved: age, BMI, the American Society of Anesthesiologists (ASA) physical status class, history of hypertension, current or recent smoking, motion sickness, or previous PONV.
      The presence of perioperative complications during the first postoperative month was graded according to the modified Clavien–Dindo (CD) classification.
      • Clavien P.A.
      • Barkun J.
      • de Oliveira M.L.
      • et al.
      The Clavien-Dindo classification of surgical complications.
      The CD grades were converted to six categories for use in the statistical analysis, with grades 1, 2, 3a, 3b, 4a, and 4b converted to categories 1, 2, 3, 4, 5, and 6, respectively.

      Apfel C C., Laara E., Koivuranta M., Greim C., Roewer N. A simplified risk score for predicting postoperative nausea and vomiting anesthesiology 1999; 91:693–700.

      LOS was calculated as the number of nights spent in the hospital postoperatively.
      DIEP flap failure was defined as the need for a revision procedure with a failure to preserve the initial flap. We also recorded the total volume of intraoperative fluids, duration of surgery, and laterality of the DIEP flap (bilateral or unilateral). The simplified Apfel score

      Apfel C C., Laara E., Koivuranta M., Greim C., Roewer N. A simplified risk score for predicting postoperative nausea and vomiting anesthesiology 1999; 91:693–700.

      (based on nonsmoking status, female sex, history of previous PONV or motion sickness, and postoperative opioid use) was calculated for each patient, and the presence of nausea or vomiting, as well as the number of antiemetics administered before and after PONV, was recorded. We likewise recorded the maximum visual analog scale (VAS) and opioid usage during the first 24 h postoperatively. Opioid usage was converted to total iv morphine equivalents as follows: 1 mg iv morphine = 1 mg iv or subcutaneous piritramide, 10 mg iv tramadol, or 2 mg sublingual oxycodone.
      • McPherson M.L.
      Demystifying opioid conversion calculations: a guide for effective dosing.
      Postoperative flap skin temperature was measured every hour for the first 24 h after surgery and compared to the temperature of the surrounding skin (reference temperature).

       Statistical analysis

      The demographic data and clinical outcome parameters are expressed as the means and standard deviations, and were either analyzed using the Kolmogorov–Smirnov test or expressed as a percentage of cases and analyzed using the χ2 test. A multiple linear or logistic regression analysis was used to determine the independent effects of different continuous and categorical variables. Variables with an autocorrelation were removed. Variables were chosen based on a proven or assumed effect with a maximum of one predictor variable per 10 observations in the smallest group.
      • Babyak M.A.
      What you see may not be what you get: a brief, nontechnical introduction to overfitting in regression-type models.
      Statistical analyses were performed using the statistical package Stata/IC 15.1 for Mac (StataCorp, TX, USA).

      Results

       Patient characteristics

      A total of 204 patients underwent unilateral or bilateral DIEP flap surgery at our institution from January 2014 to April 2019, 73 of whom received bilateral DIEP flaps. Demographic and clinical characteristics of the OFA (n = 55) and OA (n = 149) groups are shown in Table 1. The simplified Apfel score was higher in the OA group (P < 0.001), whereas age was higher in the OFA group. The age distribution was differently distributed between both groups justifying the Kolmogorov–Smirnov test instead of the Mann–Whitney U test.
      Table 1Patient characteristics.
      CharacteristicOFAOAP-value
      (n = 55)(n = 149)
      Age (years)53.92±2.7650.32±1.750.045
      Kolmogorov–Smirnov test.
      Body mass index (kg/m2)25.01±0.8325.33±0.650.609
      Kolmogorov–Smirnov test.
      ASA physical status class1.24±0.161.13±0.061.0
      Kolmogorov–Smirnov test.
      Hypertension (%)9.16.00.444
      χ2 test. Data are presented as mean ± standard deviation except where indicated otherwise. ASA, American Society of Anesthesiologists; OA, opioid anesthesia; OFA, opioid-free anesthesia; and PONV, postoperative nausea or vomiting.
      Smoker (%)9.13.40.092
      χ2 test. Data are presented as mean ± standard deviation except where indicated otherwise. ASA, American Society of Anesthesiologists; OA, opioid anesthesia; OFA, opioid-free anesthesia; and PONV, postoperative nausea or vomiting.
      Motion sickness or previous PONV (%)10.912.80.722
      χ2 test. Data are presented as mean ± standard deviation except where indicated otherwise. ASA, American Society of Anesthesiologists; OA, opioid anesthesia; OFA, opioid-free anesthesia; and PONV, postoperative nausea or vomiting.
      Simplified Apfel score2.42±0.183.01±0.08<0.001
      Kolmogorov–Smirnov test.
      Bilateral (%)38.234.90.664
      χ2 test. Data are presented as mean ± standard deviation except where indicated otherwise. ASA, American Society of Anesthesiologists; OA, opioid anesthesia; OFA, opioid-free anesthesia; and PONV, postoperative nausea or vomiting.
      Statistically significant p values are indicated in bold.
      a Kolmogorov–Smirnov test.
      b χ2 test.Data are presented as mean ± standard deviation except where indicated otherwise.ASA, American Society of Anesthesiologists; OA, opioid anesthesia; OFA, opioid-free anesthesia; and PONV, postoperative nausea or vomiting.

       Complications

      The observed complications (grades I–V) are shown in Table 2. Grades I and II represent minor complications, whereas grades III, IV, and V represent major complications. No patients in either group experienced grade IV or V complications. The rate of minor complications was lower in the OFA group than in the OA group (P < 0.001).
      Table 2Complications categorized by the Clavien–Dindo classification.
      ComplicationOFA (n = 55)OA (n = 149)P-value (χ2)
      CD I967
      CD II09<0.001
      CD I or II (minor complications)9 (17.9%)76 (51.4%)<0.001
      0
      CD IIIa01
      CD IIIb18
      CD IVa00
      CD IVb00
      CD V00
      CD III to V (major complications)1 (1.8%)9 (6.1%)0.205
      Statistically significant p values are indicated in bold.
      OFA, opioid-free anesthesia; OA, opioid anesthesia; and CD, Clavien–Dindo classification.
      Nine surgical revisions were performed under general anesthesia (CD IIIb complications): one in the OFA group and eight in the OA group. The single patient in the OFA group required revision for combined arterial and venous thromboses. In the OA group, revision was required for arterial thrombosis in one patient, venous thrombosis in four patients, combined arterial and venous thromboses in one patient, and nonperfusion-related indications in two patients. Three patients in the OA group developed flap failure that required additional reconstruction using a non-free flap; one of these patients had undergone bilateral DIEP flap surgery and the other two patients had received a unilateral DIEP flap. One patient in the OA group underwent wound revision under local anesthesia (CD IIIa). Five revisions were performed out of 131 unilateral procedures (3.8%) and 5 revisions were performed out of 73 bilateral procedures (6.8%); these differences were not statistically significant (P = 0.336). The number of minor complications was not different between the bilateral and unilateral procedures (41.1% vs. 42.7%, respectively).

       Clinical outcomes

      The clinical outcomes of both groups are shown in Table 3. There were no significant differences between groups for surgery duration, first reference skin temperature in the post-anesthesia care unit (PACU), incidence of postoperative vomiting, rate of flap failure or flap thrombosis, rate of reoperation in the first postoperative week, and LOS.
      Table 3Clinical outcomes.
      OutcomeOFAOAP-value
      (n = 55)(n = 149)
      Number of prophylactic antiemetic drugs1.13±0.181.94±0.18<0.001
      Kolmogorov–Smirnov test.
      Postoperative nausea (%)12.743.6<0.001
      χ2 test. Data are presented as mean ± standard deviation except where indicated otherwise. OFA, opioid-free anesthesia; OA, opioid anesthesia; VAS, visual analog pain score; PACU, postanesthesia care unit; and LOS, length of hospital stay.
      Number of therapeutic antiemetic drugs0.14±0.120.57±0.12<0.001
      Kolmogorov–Smirnov test.
      Postoperative vomiting (%)1.86.70.170
      χ2 test. Data are presented as mean ± standard deviation except where indicated otherwise. OFA, opioid-free anesthesia; OA, opioid anesthesia; VAS, visual analog pain score; PACU, postanesthesia care unit; and LOS, length of hospital stay.
      Maximum postoperative VAS1.87±0.654.94±0.54<0.001
      Kolmogorov–Smirnov test.
      Total morphine equivalents postoperatively (mg)1.95±0.8410.34±1.40<0.001
      Kolmogorov–Smirnov test.
      Intraoperative fluids (mL/kg/h)1.70±0.305.45±0.440.001
      Kolmogorov–Smirnov test.
      Duration of surgery (h)6.42±0.956.73±0.320.854
      Kolmogorov–Smirnov test.
      First reference skin temperature in PACU (°C)34.96±0.4134.90±0.350.186
      Kolmogorov–Smirnov test.
      Skin temperature difference (flap – reference) (°C)1.04±0.251.41±0.210.725
      Kolmogorov–Smirnov test.
      Reoperation during the first week (n, %)1 (1.8)9 (6.1)0.205
      χ2 test. Data are presented as mean ± standard deviation except where indicated otherwise. OFA, opioid-free anesthesia; OA, opioid anesthesia; VAS, visual analog pain score; PACU, postanesthesia care unit; and LOS, length of hospital stay.
      Flap thrombosis (arterial or venous) (n, %)1 (1.8)5 (3.4)0.389
      χ2 test. Data are presented as mean ± standard deviation except where indicated otherwise. OFA, opioid-free anesthesia; OA, opioid anesthesia; VAS, visual analog pain score; PACU, postanesthesia care unit; and LOS, length of hospital stay.
      Flap failure (n, %)0 (0)3 (2)0.283
      χ2 test. Data are presented as mean ± standard deviation except where indicated otherwise. OFA, opioid-free anesthesia; OA, opioid anesthesia; VAS, visual analog pain score; PACU, postanesthesia care unit; and LOS, length of hospital stay.
      No opioid usage postoperatively (%)60.012.7<0.001
      χ2 test. Data are presented as mean ± standard deviation except where indicated otherwise. OFA, opioid-free anesthesia; OA, opioid anesthesia; VAS, visual analog pain score; PACU, postanesthesia care unit; and LOS, length of hospital stay.
      Hospital LOS (days)6.82±0.267.52±0.280.134
      Kolmogorov–Smirnov test.
      Statistically significant p values are indicated in bold.
      a Kolmogorov–Smirnov test.
      b χ2 test.Data are presented as mean ± standard deviation except where indicated otherwise.OFA, opioid-free anesthesia; OA, opioid anesthesia; VAS, visual analog pain score; PACU, postanesthesia care unit; and LOS, length of hospital stay.
      The volume of intraoperative fluids was lower in the OFA group (1.70 ± 0.30 mL/kg/h) than in the OA group (5.45 ± 0.44 mL/kg/h). Patients in the OFA group received a significantly lower number of prophylactic, as well as therapeutic, antiemetic drugs, and they had a lower incidence of postoperative nausea. They also received a lower quantity of opioids (morphine equivalents) and had a lower maximum VAS score in the first 24 h postoperatively; 40% of patients receiving OFA required opioids postoperatively in contrast to 87.3% of patients in the OA group.
      When considering patients who received postoperative morphine, PONV occurred in 22.7% of patients in the OFA group and in 47.3% of those in the OA group (P = 0.034). Patients with no history of motion sickness or PONV did not receive prophylactic antiemetics if they were in the OFA group; their rate of PONV was 6.25%. If an opioid was administered after OFA in patients with no history of PONV or motion sickness, the incidence of PONV rose to 16.67% (P = 0.042). One or more prophylactic antiemetic drugs were always administered to patients who received OA. Even with the administration of prophylactic antiemetics in patients with no history of PONV, the incidence of PONV was 43.75% in OA patients who received no further opioids in the PACU and 46.39% in those who received additional opioids in the PACU (P = 0.419).
      Most patients (all except 2) with a history of PONV or motion sickness received an opioid postoperatively; their incidence of PONV were 50.00% and 60.87% after OFA and OA, respectively. (P = 0.659).

       Regression analysis results

      Table 4 shows the results of analyses evaluating the association between various factors and PONV (logistic regression analysis), CD ranking (linear regression analysis), and LOS (linear regression analysis).
      Table 4Linear regression analysis for postoperative nausea and vomiting, Clavien–Dindo complications, and length of hospital stay.
      FactorPONV
      Results of the logistic regression analysis.
      CD ranking
      Results of the linear regression analysis. PONV, postoperative nausea and vomiting; CD, Clavien–Dindo classification; LOS, length of hospital stay; coef, coefficient; OFA, opioid-free anesthesia; OA, opioid anesthesia; unilat, unilateral; and bilat, bilateral.
      LOS
      Results of the linear regression analysis. PONV, postoperative nausea and vomiting; CD, Clavien–Dindo classification; LOS, length of hospital stay; coef, coefficient; OFA, opioid-free anesthesia; OA, opioid anesthesia; unilat, unilateral; and bilat, bilateral.
      B coefP-valueB coefP-valueB coefP-value
      Type of anesthesia (OFA=1 and OA=0)–1.821<0.001–0.5360.001–0.789<0.001
      Surgery laterality (unilat=1 and bilat=2)−0.0310.8170.2380.287
      Age–0.0330.032–0.0150.0180.0260.012
      Body mass index0.0070.6790.117<0.001
      Smoker–0.7750.495
      Previous PONV0.9180.0500.0890.646−0.1120.724
      Number of prophylactic antiemetics–0.0610.707
      Statistically significant p values are indicated in bold.
      a Results of the logistic regression analysis.
      b Results of the linear regression analysis.PONV, postoperative nausea and vomiting; CD, Clavien–Dindo classification; LOS, length of hospital stay; coef, coefficient; OFA, opioid-free anesthesia; OA, opioid anesthesia; unilat, unilateral; and bilat, bilateral.
      The OFA group had a lower incidence of PONV, lower CD ranking, and shorter LOS. Younger patients had a higher incidence of PONV, higher CD ranking, and shorter LOS. Higher BMI was associated with a longer LOS.
      Table 5 shows the results of linear regression analysis evaluating the maximum postoperative VAS pain score and postoperative morphine consumption. OFA was associated with a lower VAS pain score and lower postoperative morphine consumption.
      Table 5Linear regression analysis for maximum VAS score and morphine consumption postoperatively.
      FactorMaximum VAS
      Results of the linear regression analysis. coef, coefficient; OFA, opioid-free anesthesia; OA, opioid anesthesia; unilat, unilateral; and bilat, bilateral.
      Postoperative morphine
      Results of the linear regression analysis. coef, coefficient; OFA, opioid-free anesthesia; OA, opioid anesthesia; unilat, unilateral; and bilat, bilateral.
      B coefP-valueB coefP-value
      Type of anesthesia (OFA=1 and OA=0)–2.925<0.001–8.114<0.001
      Surgery laterality (unilat=1 and bilat=2)–0.3520.436–0.1970.855
      Age–0.0320.112−0.0660.173
      Body mass index–0.0720.2290.0120.931
      Statistically significant p values are indicated in bold.
      a Results of the linear regression analysis.coef, coefficient; OFA, opioid-free anesthesia; OA, opioid anesthesia; unilat, unilateral; and bilat, bilateral.
      Table 6 shows the results of linear regression analysis evaluating the difference in skin temperature between the flap and surrounding (reference) skin. OA, high BMI, and a high first reference skin temperature in the PACU were associated with a greater skin temperature difference.
      Table 6Linear regression analysis for skin temperature difference between the flap and surrounding skin.
      FactorSkin temperature difference
      Results of the linear regression analysis.
      (flap – reference
      The reference was an area of skin surrounding the flap. coef, coefficient; OFA, opioid-free anesthesia; OA, opioid anesthesia; unilat, unilateral; bilat, bilateral; VAS, visual analog pain score; and PACU, post anesthesia care unit.
      )
      B coefP-value
      Type of anesthesia (OFA=1 and OA=0)–0.5860.007
      Age–0.0160.069
      Body mass index0.0580.023
      Maximum postoperative VAS–0.0550.096
      First reference skin temperature in PACU0.1430.003
      Statistically significant p values are indicated in bold.
      a Results of the linear regression analysis.
      b The reference was an area of skin surrounding the flap.coef, coefficient; OFA, opioid-free anesthesia; OA, opioid anesthesia; unilat, unilateral; bilat, bilateral; VAS, visual analog pain score; and PACU, post anesthesia care unit.

      Discussion

      In this study, we used a retrospective cohort to include every patient who underwent DIEP flap surgery over a 4-year period. OFA was used in approximately 25% of our patients, distributed equally over the study years. This usage was primarily based on the random availability of an anesthesiologist experienced with OFA.
      The limitations of this study include the retrospective analysis and overall low incidence of major complications and flap failures as compared to previous cohort studies. Intraoperative perfusion pressures and the total dose of vasoconstrictors were not recorded in sufficient detail to allow for analyses. Analgesics are not required intraoperatively but those that block orthosympathetic, metabolic, and inflammatory reactions or induce reflex depression, as Gray explained already in 1960,
      • Gray C.
      A reassessment of the signs and levels of anaesthesia.
      are required instead. Sufficient sympatholysis was verified by using ANI monitoring (MDoloris, Lille, France) in OFA patients and if the nociceptive reaction was for more than 15 min below 50%, extra dexmedetomidine and lidocaine was given.
      Previous studies in bariatric surgery reported that OFA
      • Mulier J.
      • Dillemans B
      Anesthetic factors affecting outcome after bariatric surgery, a retrospective leveled regression analysis.
      and ERAS
      • Sharif-Askary B.
      • Hompe E.
      • Broadwater G.
      • Anolik R.
      • Hollenbeck S.T
      The effect of enhanced recovery after surgery pathway implementation on abdominal-based microvascular breast reconstruction.
      reduced postoperative opioid use by 75% and improved the outcome. In the current study, OFA was associated with a more than 80% reduction in opioid consumption in the first 24 h postoperatively.
      The Clavien–Dindo classification (CD) was developed by Clavien in 1992 for cholecystectomy and modified by Dindo in 2004.
      • Dindo D.
      • Demartines N.
      • Clavien P.A
      Classification of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey.
      Recently, it has been used as an objective tool for grading morbidity after free-flap surgery.
      • Miyamoto S.
      • Nakao J.
      • Higashino T.
      • et al.
      Clavien–Dindo classification for grading complications after total pharyngolaryngectomy and free jejunum transfer.
      CD found a difference for minor complications (CD I and II), but not for major complications. The rates of flap thrombosis and flap failure were lower in this study than in previous reports.
      • Allen J.
      The evolution of perforator flap breast reconstruction: twenty years after the first DIEP flap.
      The linear regression analysis of the CD ranking showed more complications for OA and younger patients.
      Patients in our OFA group received less intraoperative fluids (1.70 mL/kg/h) than those in the OA group (5.45 mL/kg/h). Nelson et al. reported that less than 7 mL/kg/h of intraoperative fluids increased the risk of postoperative thrombosis above their average rate of 4%.
      • Nelson J.A.
      • Fischer J.P.
      • Grover R.
      • et al.
      Intraoperative perfusion management impacts postoperative outcomes: an analysis of 682 autologous breast reconstruction patients.
      Our rate of thrombosis was lower in OFA and OA (1.8% vs. 3.4%). Dexmedetomidine causes direct vasoconstriction,
      • Talke P.
      • Lobo E.
      • Brown R
      Systemically administered alpha2-agonist-induced peripheral vasoconstriction in humans.
      but its indirect sympatholytic effects are more pronounced, which results in vasodilation. Free flaps are not innervated and might therefore elicit a hypothetical steal effect. Nunes did not find an induction of vasoconstriction in denervated flaps with the use of dexmedetomidine,
      • Nunes S.
      • Berg L.
      • Raittinen L.P.
      • et al.
      Deep sedation with dexmedetomidine in a porcine model does not compromise the viability of free microvascular flap as depicted by microdialysis and tissue oxygen tension.
      and we could not confirm the steal effect either, suggesting the importance of reducing the levels of circulating catecholamines. ERAS might increase perioperative hypotension and bradycardia,
      • Anolik R.A.
      • Sharif-Askary B.
      • Hompe E.
      • et al.
      Occurrence of symptomatic hypotension in patients undergoing breast free flaps; is ERAS to blame?.
      which requires additional vasoconstriction. Dexmedetomidine and lidocaine have also been reported to reduce the activation of coagulation, and might therefore be more important for the prevention of thrombosis.
      • Zhang Y.
      • Liu M.
      • Ynag Y.
      • Cao J.
      • Mi W
      Dexmedetomidine exerts a protective effect on ischemia-reperfusion injury after hepatectomy: a prospective, randomized, controlled study.
      ,
      • Ganter M.T.
      • Hofer C.K.
      • Spahn D.R.
      • et al.
      The effect of clonidine on perioperative blood coagulation.
      One of our patients who received 6.1 mL/kg/h of fluids intraoperatively developed fluid overload, with delayed wound healing and prolonged hospitalization. This outcome appears contrary to Nelson's recommendation to give 8 mL/kg/h.
      • McPherson M.L.
      Demystifying opioid conversion calculations: a guide for effective dosing.
      The judicious use of vasoconstrictors to maintain perfusion pressure may be superior to high fluid intake to avoid wound edema.
      • Anker A.M.
      • Prantl L.
      • Strauss C.
      • et al.
      Assessment of DIEP flap perfusion with intraoperative indocyanine green fluorescence imaging in vasopressor-dominated hemodynamic support versus liberal fluid administration: a randomized controlled trial with breast cancer patients.
      A clinical trial
      • Anker A.M.
      • Prantl L.
      • Strauss C.
      • et al.
      Vasopressor support vs. liberal fluid administration in deep inferior epigastric perforator (DIEP) free flap breast reconstruction – a randomized controlled trial.
      found no difference in flap failure or LOS between 2 and 8 mL/kg/h of intraoperative fluids either, when norepinephrine was used to maintain the systolic blood pressure above 100 mmHg. The ERAS pathway recommends restrictive fluid therapy, with the administration of additional fluids when the pulse pressure variation exceeds 20%; these recommendations were followed in our OFA patients.
      • Forget P.
      • Lois F.
      • De Kock M
      Goal-directed fluid management based on the pulse oximeter–derived pleth variability index reduces lactate levels and improves fluid management.
      The difference in skin temperature between free flap and surrounding skin might be a measurement of quality of perfusion. No difference was found between both groups except in the regression analysis.
      Kroll et al.
      • Kroll S.S.
      • Sharma S.
      • Koutz C.
      • et al.
      Postoperative morphine requirements of free TRAM and DIEP flaps.
      reported the use of 0.74 mg/kg morphine equivalents in the first 24 h after DIEP flap surgery, which was substantially higher than the doses in both our OA and OFA groups (0.15 mg/kg and 0.03 mg/kg, respectively). Our findings show the importance of reducing intraoperative opioids to zero. The Apfel score differed between groups, because fewer patients required opioids after OFA.
      PONV is common after DIEP flap breast reconstruction surgery, which occurs in up to 76% of patients
      • Manahan M.A.
      • Basdag B.
      • Kalmar C.L.
      • et al.
      Risk of severe and refractory postoperative nausea and vomiting in patients undergoing DIEP flap breast reconstruction.
      . PONV begins soon after awakening and persists for several days, notwithstanding the use of multiple rescue antiemetics. Although OFA itself reduces PONV, opioid use after OFA still increases PONV. Patients with previous PONV, more frequently required a small dose of opioids for pain relief in the PACU, even after OFA, and had a high rate of PONV despite the use of more prophylactic antiemetics. Apfel measured that TIVA as compared to inhalation reduced PONV because of the antiemetic effects of propofol from 38% to 31%.
      • Apfel C.C.
      • Korttila K.
      • Abdalla M.
      • et al.
      A factorial trial of six interventions for the prevention of postoperative nausea and vomiting.
      Three of the seven cases receiving TIVA without any inhalation still had PONV. All DIEP flap patients in this study received propofol in a continuous infusion. The incidence of 44% in the OA group is still higher than the 13% in the OFA group.
      Wade et al. recently reported a higher revision rate after bilateral DIEP flap surgery (23%) than after a unilateral DIEP flap procedure (11%)
      • Wade R.G.
      • Razzano S.
      • Sassoon E.M.
      • et al.
      Complications in DIEP flap breast reconstruction after mastectomy for breast cancer: a prospective cohort study comparing unilateral versus bilateral reconstructions.
      . We likewise observed twice the number of revisions after bilateral surgery, albeit at a much lower frequency (6.8% vs. 3.8%).
      LOS was not shorter in the OFA group except when other factors were included in a regression.
      With a larger database, one can discuss the value of adding more variables in the regression analysis, but this would be based more on assumptions than on a proven impact.
      In summary, we observed no differences in major complications between OFA combined with goal-directed fluid therapy and OA for DIEP flap surgery. OFA and goal-directed fluid therapy were associated in a regression analysis with less minor complications, less pain and a reduced rate of PONV, less difference in skin temperature between the flap and surrounding skin, lower postoperative opioid use, and shorter hospital LOS.

      Declaration of Competing Interest

      Harold Mulier: None.
      Bob De Frene: None.
      Lara Benmeridja: None.
      Florian Vanhoorebeeck: None.
      Bruno Denis: None.
      Bob Casaer: None.
      Fabrice Rogge: None.
      Kris Leleu: None.
      Jan Mulier: Personal fees, department grants, financial, and nonfinancial support to organize scientific meetings from MSD Merck, GE Healthcare, and Johnson & Johnson outside the submitted work.

      Acknowledgments

      We would like to thank the members of the Department of Anesthesia, Reanimation, and Intensive Care, AZ Sint Jan Brugge, for their support of this study and for demonstrating the OFA technique to visitors from around the world. We would also like to thank members of the Department of Plastic and Reconstructive Surgery, AZ Sint Jan Brugge, for their collaboration in the search for improved patient care.

      Financial disclosures

      None for this submitted work. No support whatsoever was obtained to perform this study.
      Clinical Trial Registry and Number: Clinicaltrials.gov, NCT 03202134.

      References

        • Wade R.G.
        • Razzano S.
        • Sassoon E.M.
        • et al.
        Complications in DIEP flap breast reconstruction after mastectomy for breast cancer: a prospective cohort study comparing unilateral versus bilateral reconstructions.
        Ann Surg Oncol. 2017; 24: 1465-1474
        • Siegel R.L.
        • Miller K.D.
        • Jemal A
        Cancer statistics.
        CA Cancer J Clin. 2019; 69: 7-34
        • Blondeel P.N
        One hundred free DIEP flap breast reconstructions: a personal experience.
        Br J Plast Surg. 1999; 52: 104-111
        • Allen J.
        The evolution of perforator flap breast reconstruction: twenty years after the first DIEP flap.
        Reconstr Microsurg. 2014; 30: 121-126
        • Batdorf N.J.
        • Lemaine V.
        • Lovely J.K.
        • et al.
        Enhanced recovery after surgery in microvascular breast reconstruction.
        J Plast Reconstr Aesthet Surg. 2015; 68: 395-402
        • Mulier J.
        • Wouters R.
        • Dillemans B.
        • De Kock M
        A randomized controlled, double-blind trial evaluating the effect of opioid-free versus opioid general anaesthesia on post-operative pain and discomfort measured by the QoR-40.
        J Clin Anesth Pain Med. 2018; 2: 015
        • Manahan M.A.
        • Basdag B.
        • Kalmar C.L.
        • et al.
        Risk of severe and refractory postoperative nausea and vomiting in patients undergoing DIEP flap breast reconstruction.
        Microsurgery. 2014; 34: 112-121
        • Hagau N.
        • Longrois D.
        Anesthesia for free vascularized tissue transfer.
        Microsurgery. 2009; 29: 161-167
      1. Li, B.,Li Y., Tian S.et al. Anti-inflammatory effects of perioperative dexmedetomidineadministered as an adjunct to general anesthesia: a meta-analysis. Sci. Rep. 5, 12342.

        • Byon H.
        • Ok S.
        • Lee S.H.
        • et al.
        Dexmedetomidine inhibits phenylephrine-induced contractions via alpha-1 adrenoceptor blockade and nitric oxide release in isolated rat aortae.
        Int J Med Sci. 2017; 14: 143-149
        • Cassuto J.
        • Sinclair R.
        • Bonderovic M.
        Anti-inflammatory properties of local anesthetics and their present and potential clinical implications.
        Acta Anaesthesiol Scand. 2006; 50: 265-282
        • Gherardini G.
        • Gurlek A.
        • Cromeens D.
        • et al.
        Drug-induced vasodilation: in vitro and in vivo study on the effects of lidocaine and papaverine on rabbit carotid artery.
        Microsurgery. 1998; 18: 90-96
        • Rothenberger J.
        • Amr A.
        • Schiefer J.
        • Schaller H.
        • Rahmanian-Schwarz A
        A quantitative analysis of the venous outflow of the deep inferior epigastric flap (DIEP) based on the perforator veins and the efficiency of superficial inferior epigastric vein (SIEV) supercharging.
        J Plast Reconstr Aesthet Surg. 2013; 66: 67-72
        • Gazi M.
        • Abitagaoglu S.
        • Turan G.
        • et al.
        Evaluation of the effects of dexmedetomidine and remifentanil on pain with the analgesia nociception index in the perioperative period in hysteroscopies under general anesthesia. A randomized prospective study.
        Saudi Med J. 2018; 39: 1017-1022
        • Clavien P.A.
        • Barkun J.
        • de Oliveira M.L.
        • et al.
        The Clavien-Dindo classification of surgical complications.
        Ann Surg. 2009; 250: 187-196
      2. Apfel C C., Laara E., Koivuranta M., Greim C., Roewer N. A simplified risk score for predicting postoperative nausea and vomiting anesthesiology 1999; 91:693–700.

        • McPherson M.L.
        Demystifying opioid conversion calculations: a guide for effective dosing.
        American Society of Health-System Pharmacists, Inc, 2009 (ISBN 978-1-58528-198-5)
        • Babyak M.A.
        What you see may not be what you get: a brief, nontechnical introduction to overfitting in regression-type models.
        Psychosom Med. 2004; 66: 411-421
        • Gray C.
        A reassessment of the signs and levels of anaesthesia.
        Ir J Med Sci. 1960; 419: 499-508
        • Mulier J.
        • Dillemans B
        Anesthetic factors affecting outcome after bariatric surgery, a retrospective leveled regression analysis.
        Obes Surg. 2019; 29: 1841-1850
        • Sharif-Askary B.
        • Hompe E.
        • Broadwater G.
        • Anolik R.
        • Hollenbeck S.T
        The effect of enhanced recovery after surgery pathway implementation on abdominal-based microvascular breast reconstruction.
        J Surg Res. 2019; 242: 276-285
        • Dindo D.
        • Demartines N.
        • Clavien P.A
        Classification of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey.
        Ann Surg. 2004; 240: 205-213
        • Miyamoto S.
        • Nakao J.
        • Higashino T.
        • et al.
        Clavien–Dindo classification for grading complications after total pharyngolaryngectomy and free jejunum transfer.
        PLoS ONE. 2019; 14e0222570
        • Nelson J.A.
        • Fischer J.P.
        • Grover R.
        • et al.
        Intraoperative perfusion management impacts postoperative outcomes: an analysis of 682 autologous breast reconstruction patients.
        J Plast Reconstr Aesthet Surg. 2016; 68: 175e183
        • Talke P.
        • Lobo E.
        • Brown R
        Systemically administered alpha2-agonist-induced peripheral vasoconstriction in humans.
        Anesthesiology. 2003; 99: 65-70
        • Nunes S.
        • Berg L.
        • Raittinen L.P.
        • et al.
        Deep sedation with dexmedetomidine in a porcine model does not compromise the viability of free microvascular flap as depicted by microdialysis and tissue oxygen tension.
        Anesth Analg. 2007; 105: 666-672
        • Anolik R.A.
        • Sharif-Askary B.
        • Hompe E.
        • et al.
        Occurrence of symptomatic hypotension in patients undergoing breast free flaps; is ERAS to blame?.
        Plast Reconstr Surg. 2020; 145: 606-616
        • Zhang Y.
        • Liu M.
        • Ynag Y.
        • Cao J.
        • Mi W
        Dexmedetomidine exerts a protective effect on ischemia-reperfusion injury after hepatectomy: a prospective, randomized, controlled study.
        J Clin Anesth. 2019; https://doi.org/10.1016/j.jclinane.2019.109631
        • Ganter M.T.
        • Hofer C.K.
        • Spahn D.R.
        • et al.
        The effect of clonidine on perioperative blood coagulation.
        J Clin Anesth. 2005; 17: 456-462
        • Anker A.M.
        • Prantl L.
        • Strauss C.
        • et al.
        Assessment of DIEP flap perfusion with intraoperative indocyanine green fluorescence imaging in vasopressor-dominated hemodynamic support versus liberal fluid administration: a randomized controlled trial with breast cancer patients.
        Ann Surg Oncol. 2020; 27: 399-406
        • Anker A.M.
        • Prantl L.
        • Strauss C.
        • et al.
        Vasopressor support vs. liberal fluid administration in deep inferior epigastric perforator (DIEP) free flap breast reconstruction – a randomized controlled trial.
        Clin Hemorheol Microcirc. 2018; 69: 37-44
        • Forget P.
        • Lois F.
        • De Kock M
        Goal-directed fluid management based on the pulse oximeter–derived pleth variability index reduces lactate levels and improves fluid management.
        Anesth Analg. 2010; 111: 910-914
        • Kroll S.S.
        • Sharma S.
        • Koutz C.
        • et al.
        Postoperative morphine requirements of free TRAM and DIEP flaps.
        Plast Reconstr Surg. 2001; 107: 338-341
        • Apfel C.C.
        • Korttila K.
        • Abdalla M.
        • et al.
        A factorial trial of six interventions for the prevention of postoperative nausea and vomiting.
        N Engl J Med. 2004; 350: 2441-2451

      Linked Article