Abstract
BACKGROUND:
Perioperatively, patients’ hemodynamics are modulated predominantly by intravenous fluid administration and vasoactive pharmacological support. Vasopressor agents are suspected to be detrimental on free flap survival by the cause of vasoconstriction of the pedicle with consecutive reduced overall flap perfusion and by aggravation of flap dissection.
OBJECTIVE:
A novel, standardized fluid restrictive perioperative hemodynamic management was assessed for its feasibility in clinical practice in free flap patients undergoing breast reconstruction.
METHODS:
Patients were randomized to two perioperative regimens with different fluid and vasopressor limits. The primary endpoint regarded flap survival. Secondary endpoints included surgery times, time of patient ambulation and length of hospital stay.
RESULTS:
There was one total flap failure with liberal fluid administration (LFA). No total or partial flap failure was noted in the fluid restrictive regimen with norepinephrine administration up to 0.04μg/kg/min (FRV). No delay regarding operation time (p = 0.217), patient mobilization (p = 0.550) or hospital discharge (p = 0.662) was registered in the FRV study subpopulation compared to LFA.
CONCLUSIONS:
The results of this prospective interventional trial could not detect any negative impact of vasopressors, neither for the primary endpoint of flap survival nor for the overall patient outcome. The fear of vasopressor associated flap complications has led to a traditional liberal fluid administration, which failed to demonstrate any benefits when compared to a fluid restrictive vasopressor strategy.
Keywords
Introduction
Microsurgical tissue transfer has become a standard treatment for soft tissue as well as bone reconstruction over the past two decades. In the context of autologous breast reconstruction, the deep inferior epigastric perforator (DIEP) flap has evolved into a first-choice procedure of tissue transfer [1, 2]. The DIEP flap for breast reconstruction consists of the free transfer of abdominal pannus to reconstruct the breast mound during several hours of general anesthesia.
It is well-known, that regulation of blood pressure is vital to maintain optimal end-organ oxygen delivery [3, 4]. Perioperatively, patient hemodynamics are modulated predominantly by intravenous fluid administration and vasoactive pharmacological support. Yet, the extensive use of vasopressor agents (e.g., norepinephrine) for blood pressure management during microvascular procedures is still a subject of controversy [5, 6]. Vasopressor agents are suspected to be detrimental on flap survival by the cause of vasoconstriction of the pedicle with consecutive reduced overall flap perfusion and by aggravation of flap harvest [7]. Du et al. could show that an increase of the mean arterial pressure from <60 mmHg to 60–90 mmHg under norepinephrine did not lead to an improvement of the microcirculation [8]. The surgeon’s fear of vasopressor associated flap complications has therefore led to a traditional liberal fluid administration (LFA) in order to avoid higher dosages of vasoactive substances. However, this strategy of increased fluid intake may frequently result in positive fluid balances with corresponding complications such as interstitial fluid accumulation, associated decrease in flap oxygenation and possible impacts on blood coagulation [9]. In fact, emerging evidence suggests that excessive perioperative fluid administration may have negative impacts on multiple aspects of recovery after major surgery including flap failure [6, 9–18].
Nevertheless, to date no trial compared LFA to fluid restriction and vasopressor support (FRV) in patients undergoing DIEP flap breast reconstruction.
In this prospective study two perioperative regimens with different fluid and vasopressor limits for the regulation of systemic blood pressure were investigated regarding their possible detrimental effects on flap survival in patients undergoing autologous breast reconstruction with the DIEP flap.
Material and methods
Study design
The study is an interventional, randomized clinical trial. Study start date was March 13, 2017. The last patient was enrolled on October 10, 2017.
The patients were recruited during a pre-assessment visit at the study center, where written and oral information was provided. Upon written informed consent participants were randomized into one of the two study arms by manual randomization (LFA vs. FVC). Participants as well as surgeons were blinded for treatment allocation. The investigator was not blinded.
Surgical technique
Preoperatively, the fusiform flap is marked with the superior incision placed just above the umbilicus, extending laterally and finishing adjacent to the anterior superior iliac spines. The lower aspect of the incision is placed transversely at the level of the caesarian section, gently curving lateral, meeting the upper marking [19].
In the operating room the patient is placed in the supine position. Two surgical teams simultaneously perform abdominal flap harvest and preparation of the thoracic recipient vessels following mastectomy in immediate breast reconstruction. The flap is harvested on a single perforator of the medial or lateral row as previously described [19]. The internal mammary artery and vein are consistently chosen as recipient vessels. Conventional end-to-end arterial anastomosis is performed with interrupted 8-0 monofilament sutures. A coupler device (Synovis Micro Companies Alliance, Birmingham, AL, USA) is used for venous anastomosis (diameter ranging from 2.0 to 3.5 mm).
Intermittent pneumatic compression is applied for the lower limbs to improve venous circulation and prevent deep venous thrombosis during the procedure.
Anesthetics and intervention
The study compared two regimens to maintain systemic circulation with different fluid and vasopressor limits as presented in Table 1.
Intervention variables
Intervention variables
(FRV – fluid restriction and vasopressor support, LFA – liberal fluid administration).
Throughout patients received balanced crystalloid solutions (Jonosteril, Fresenius Kabi, Bad Homburg, Germany) via central venous access. No colloids were administered. The vasopressor agent used was norepinephrine (Arterenol, Sanofi-Aventis, Frankfurt, Germany).
In both study subpopulations, anesthesia was induced with propofol 1.5–2.0 mg/ kg (Braun, Melsungen, Germany) and sufentanil 0.25–0.5μg/ kg (hameln pharma plus GmbH, Hameln, Germany). Orotracheal intubation was facilitated with atracurium 0.3–0.6 mg/ kg (Hikma Farmacêutica, Terrugem, Portugal). Anesthesia was maintained using sevoflurane (Baxter Dräger, Vapor 2000) and intermittent administration of sufentanil.
The intraoperative systolic target blood pressure was above 100 mmHg. An arterial catheter was placed into the radial artery to ensure continuous blood pressure assessment.
A Foley catheter was used for urinary drainage control and simultaneous monitoring of core body temperature to prevent hypothermia.
The primary endpoint was defined as total flap failure until day 5 postoperative. Secondary outcome variables were revision surgery (e.g., due to thromboembolic complication or hematoma), surgery times, time of patient ambulation and length of hospital stay.
Inclusion and exclusion criteria
Voluntarily participation was offered to all patients who desired autologous breast reconstruction with the DIEP flap at the recruiting center. Immediate as well as secondary breast reconstructions were equally enrolled in this study.
Exclusion criteria were denial of written consent, high grade obesity (BMI >35 kg/m2) and contraindications for one of the applied regimens.
Statistical analysis
Data is expressed as arithmetic mean±standard deviation (SD) as well as median values with interquartile range (IQR). Gaussian distribution of data was confirmed by Shapiro-Wilk-test, p > 0.05.
Statistical significances were analyzed by unpaired Student’s t-test. P-values of <0.05 were considered significant. Statistical analysis was performed using SPSS version 24.0.0.0 software (SPSS Inc., USA).
Ethical approval and trial registration
The study was conducted upon approval and in accordance with the University of Regensburg institutional ethics committee (Reference no. 16-101-0293).
The study has been registered on the international ClinicalTrials.gov database (Reference no. NCT03118024).
Results
Patient variables
Twelve patients were enrolled in the study (n = 12). Five participants were randomly assigned to the FRV group (n1 = 6), 6 patients were enrolled in the LFA group (n2 = 6). One patient was excluded from the study retrospectively, as the intraoperative perforator quality was poor, which necessitated intraoperative conversion to a transverse rectus abdominis myocutaneous flap (TRAM) (ncorr = 11, n2corr = 5, n1corr = n1 = 6).
The study cohort comprised homogenously distributed demographics across the subpopulations as demonstrated above (Table 2).
Patient variables
Patient variables
One total flap failure was observed in the LFA study arm due to postoperative arterial occlusion. No flap loss or partial flap failure was noted in the FRV regimen.
Secondary study outcomes (Table 3)
Mean operation time accounted for 389 min (t = 389±52). With 371 min (t1 = 371±57) in the FRV group and 411 min (t2 = 411±38) with LFA, there was a trend towards faster operation time in the fluid restrictive regimen, however, values were not significantly different (p = 0.217). Immediate (n1 = 7) and secondary reconstructions (n2 = 4) were included.
Patients were mobilized as fast as 1.3 days (x1 = 1.3±0.5) postoperatively with FRV, and 1.6 days (x2 = 1.6±0.9) postoperatively in the LFA population (p = 0.550). One patient with intraoperative vasoconstrictor restriction was postoperatively monitored for 72 hours at the intermediate care unit because of polyuria.
The period of hospitalization accounted for 7.2 postoperative days (h2 = 7.2±0.4) in the LFA group and for 7.0 days (h1 = 7.0±0.9) with FRV (p = 0.662).
Secondary outcome results
Secondary outcome results
There was no anastomosis-related revision but one revision surgery for hematoma evacuation beneath the flap on postoperative day 3 to prevent venous congestion in the FRV subpopulation. In the LFA group no revision was registered until postoperative day 5.
To date there is a lack of data supporting the long-established fear of detrimental effects of vasoconstrictor use in microvascular surgery [20, 21]. However, emerging evidence supports a negative impact of extensive intraoperative fluid administration on medical and surgical outcome in free autologous tissue transfer [10–13]. Nevertheless, existing data consists of retrospective reviews only, which precludes measurements of predefined dosage limits. Additionally, previous trials compared different types of reconstructive options, which makes statistical analysis and consecutive extrapolation of this retrospective data problematic.
The DIEP flap has universally become the workhorse for autologous breast reconstruction in most institutions, thus making this flap the most frequently applied free flap among currently available free tissue transfer options [1]. Further, harvest site, recipient site, flap size and conditions for flap anastomoses are highly standardized in this particular indication. Patient demographics that require DIEP flap breast reconstruction are universally analogous, as all patients are of female sex, within their fourth or fifth decade of life, in a moderately over-weighted constitution and in an adequate health condition allowing this elective microsurgical procedure. With regard to interventional studies, this uniform cohort in combination with the standardized treatment modalities of the DIEP flap provide ideal utility for the prospective direct comparison of treatment regimens.
This study comprised a LFA regimen as traditionally applied at most centers, which was compared to a novel FRV protocol. Based on best available evidence, the FRV protocol included higher intraoperative norepinephrine doses [12, 22].
With respect to the predefined limits of vasopressor and volume administration, in this cohort both circulatory regimens allowed the constant maintenance of 100 mmHg systolic blood pressure. Most importantly, neither of the applied circulatory regimens caused hemodynamical or surgical complications perioperatively.
Nonetheless, the primary endpoint of this study was defined as the rate of flap failure until postoperative day 5. In the LFA group, one complete flap failure was detected due to arterial occlusion. According to the literature, despite progressive spread of microsurgical experience, the risk for complete failure of free microvascular flaps ranges from 5–7% [23–28]. Whether the flap failure was owed to the LFA protocol or simple coincidence is not ultimately verifiable. However, other authors identified intraoperative volume overload as a risk factor for flap failure recently [6, 10–13]. Fluid administration exceeding 7 L or >5.4 ml/kg/h (>130 ml/kg/day) during microvascular head and neck reconstruction was related to medical and surgical complications including flap failure [12, 13]. Further, a retrospective study in the context of free flap breast reconstructions determined the excessive intraoperative crystalloid infusion rate as an independent predictor for postoperative complications [10]. Regarding these reports, the observation of this study is in line with data of previous studies in this field and suggests a higher risk of flap failure for LFA regimens. Conversely, partial or complete flap losses could not be observed in the FRV study arm. Despite that, one revision surgery due to hematoma occurred. The revision was not related to the anastomosis and attributable to venous bloody oozing from the thoracic subcutaneous pocket.
Norepinephrine is suspected to impede perforator dissection as it mediates vasoconstriction via its action on α1-receptors of vascular endothelial cells [29]. Furthermore, prolonged surgery times are associated with higher incidence of postoperative complications [10]. Interestingly, surgery times did not differ significantly between study arms in this trial (p = 0.217). As two teams operated simultaneously, the consistent time limiting factor was ultimately flap dissection. Irrespective from simultaneous mastectomy, thoracic pocket and recipient vessel preparation had always been fully prepared before completion of flap harvest. Further, flap dissection was performed by one-out-of-two lead surgeons with equivalent microsurgical experience. With regard to operation time, vasoconstriction seems to have a negligible practical relevance according to our data.
Early postoperative mobilization has been identified as a paramount factor for morbidity and length of hospital stay [30]. Reduced intravascular volume is associated with increased orthostatic intolerance and may thereby be a hindrance for early patient ambulation [31]. Remarkably, lower intraoperative fluid volumes in the FRV regimen did not cause significant delay in postoperative mobilization of the patient (p = 0.550). The overall favorable cardiovascular condition of patients in this cohort might be one explanation for this observation. Further, strategies of goal-directed fluid management in other fields of surgery could not reduce the prevalence of orthostatic intolerance and emphasize the complexity of this phenomenon [31]. Although impaired cardiac output and peripheral resistance contribute to orthostatic intolerance, their precise pathogenic mechanisms in postoperative autonomic dysfunction are still under investigation [31–34]. Conversely, there was a trend towards faster mobilization with FRV (x1 = 1.2±0.4), compared to the LFA regimen (x2 = 1.5±1.0; p = 0.665). Consequently, this is highly favorable in terms of fast-track rehabilitation, encouraging patient autonomy and reconvalescence [35].
Length of hospital stay is known to affect morbidity, patient satisfaction and cost-effectiveness [36]. Consequently, rapid recovery with early discharge from hospital has become a criterion for most novel therapies. The evaluation of duration of hospital stay in this study revealed a trend towards shorter length of hospital stay of 7.0 days (h1 = 7.0±1.0) within FRV group towards 7.3 postoperative days (h2 = 7.3±0.5) in the LFA study population. Nevertheless, this result did not withstand statistical significance (p = 0.665) when subjected to statistical analysis. Still, along with faster mobilization there is a tendency towards earlier discharge in the FRV cohort.
Conclusion
Several retrospective observational studies suggested detrimental effects on flap survival in free autologous tissue transfer for the extensive intraoperative administration of fluid [10–13].
In this trial a novel, standardized fluid restrictive perioperative hemodynamic management was assessed for its feasibility in free flap patients in clinical practice. In accordance with previous data, the results of this prospective interventional trial could not detect any negative impact of vasopressors, neither for the primary endpoint of flap survival nor for the overall patient outcome. On the contrary, one complete flap failure was recorded for the LFA group. Further, fluid restriction seemed to reduce surgery time, enhance fast mobilization as well as early hospital discharge.
The fear of vasopressor associated flap complications has led to a traditional LFA, which failed to demonstrate any benefits when compared to a FRV strategy. Future studies that monitor flap perfusion during free tissue transfer procedures will hopefully further elucidate ideal vasopressor limits to define optimal conditions for microsurgery.
Footnotes
Acknowledgment
The publication is dedicated to the 70th birthday of Prof. F. Jung.