DIEP flaps at our institution

 I recently started a new position at a facility that performs a fair number of DIEP flaps after breast cancer surgery.  The surgeries are very adamant that the anesthesia provider avoid all vasopressors and give at minimum 4 to 5 liters of fluid regardless, and also use additional repaid fluid infusion to treat and prevent hypotension.  This approach is actually written down as instructions to the anesthesia providers on how to approach their patients.  In my previous practice we did flap repairs but never received direct instructions from the surgeons involved with an algorithm to treat hemodynamic perturbations.  Therefore, when I arrived here and received a "recipe" for the approach to treatment of hypotension, I was surprised.

The concept of "loading the patient with crystalloid" as indicated by our plastic surgeons derives from the idea in free flap physiology related to maintain low SVR in a relative sense with elevated cardiac output. It is clear that, when a patient is hypovolemic, the natural response in the healthy intact human is to see elevated levels of ANP and sympathetic nervouse system output resulting in vasoconstriction.  Therefore, the fear of a patient having a large sympathetic response with significant vasoconstriction and low cardiac output prompts the concept of giving large amounts of fluids to avoid this.  Unfortunately, this ignores the idea that in patients who are appropriately anesthetized, the baroreceptor response is blunted, and sympathetic response is severely mitigated.  Patients under GA find themselves in a state of low SVR although cardiac output is often not increased as desired by the free flap surgeon.  Unfortunately, large volume crystalloid may be harmful.  In an elegant review on the endothelial and endothelial glycocalyx physiology by Millford et al. [11], one can see how infusing large volumes of crystalloid may or may not remain in the intravascular space depending on the intrinsic intra capillary pressure.  As this pressure increases, any crystalloid infused will simply be "pressed" out of the intravascular space and into the interstitial space.  Here, excess fluid will then interefere with tissue oxygenation and thus degrade perfusion of the free flap intended to protect.  Furthermore, there is evidence that diluting albumin with large volume crystalloid intereferes with the endothelial glycocalyx integrity.  In theory this will allow a greater flux of fluid across the endothelial lining of capillaries resulting in greater interstitial fluid accumulation. In research that highlights this physiology, it has been shown that when healthy volunteers are hemorrhaged 900 mL, an equal amount to 2 x 's this volume of balanced crystalloid can restore normovolemia. This is an example of unique microcirculation physiology, where the intracapillary pressure is decreased by hemorrhage, thus allowing infusing crystalloid to remain in the intravascular space.  In comparison, when crystalloid in infused to induce hypervolemia, only 17% (+/-) 10% remains in the intravascular space, the rest being pushed (squeezed) into the interstitial space.  If the endothelial glycocalyx is damaged, the amount of crystalloid OR colloid infused will now be pushed into the interstitial space at lower pressures in the capillaries.  With severe endothelial glycocalyx degradation, both colloids and crystalloids will move out of the intravascular space into the interstitial space in equal volumes. The amount of fluid that is lost across the endothelial membrane is then largely a function of elevated pressure within the microvascular space AND intactness of the glycocalyx.  There is evidence to suggest that large volume crystalloid infusion can disrupt or cause breakdown of the glycocalyx. Furthermore, albumin  carries the phospholipid sphingosine 1 phosphate (S1P), which is essential in the maintenance of the endothelial glycocalyx structure. There is evidence that demonstrates that as the albumin concentration decreases thus interrupting S1P transport endothelial glycocalyx suffers significantly degradation. It has been demonstrated that in absence of albumin, 25 times less S1P is released from RBC's (its primary source). Furthermore, there is some evidence in animal models suggesting that albumin infusion can restore the glycocalyx.  However, it is clear that albumin is not likely to be able to restore  glycocalyx function directly, rather, it is the S1P that the albumin coaxes out of RBCs or the S1P in the albumin solution infused that mediates repair. On the other hand, FFP has clearly been demonstrated to be restorative to the glycocalyx structure via upregulation of endothelial glycocalyx component production.

   Thus, preserving endothelial glycocalyx structure is critical when caring patients undergoing DIEP flap (or any other free flap).  This goal is put at risk when we are asked to rapidly administer high volume crystalloid solutions (and in particular when the crystalloid consists of NS).  The damage done may be less when administering large amounts of crystalloid to patients who are hypotensive, at least while microvascular pressures remain low. However, there can be a reversal in the post operative period when baroreceptor function returns and microvascular pressures return to normal resulting in a relative hypervolemia in contradistinction to the relative HYPOvolemia produced by induction of anesthesia.

Above is a practical physiologic rationale for avoiding large volume crystalloid.  However, I wanted to see if there was good clinical evidence from real world studies that support avoidance of large volume crystalloid infusions. Karamanos et al. [1] performed a retrospective review of 126 patients undergoing DIEP flap. One group had received 5.5 mL/kg/hr vs liberal fluid group who had received 10.2 mL/kg/hr. The group who received liberal fluids (more fluid) had more wound complications (76% vs 15%). Furthermore, the flap oximetry  readings were lower in the liberal fluid group over the 24 hours following surgery (41% vs 56%). No difference in AKI between groups.  It is interesting to note that the flap oximetry readings were lower in the liberal fluid regimen group.  This is direct evidence of the extra fluid finding its way into the interstitial space thus harming tissue oxygen tissue levels. In another retrospective review on breast reconstruction surgery by Sjöberg, T et al. [2] a  moderate fluid therapy group was compared to liberal fluid therapy. It should be noted that radiotherapy was more frequent in the group that received the moderate fluid making them at higher risk for post op complications with wound healing. Neverthless, despite this, the study found that intraop and post op complications were more frequent in the liberal fluid group.  The moderate fluid therapy group included noriepi as part of the protocol to treat hypotension. Zhong T, et al. did a retrospective multivariate analysis and found that extremes of crystalloid infusions predicted post operative complications (p=0.03). This analysis carefully controlled for a multitude of different possible causes of post op complications.

On the other hand, if one follows the general theme outlined above, it is likely that low blood pressure may in some cases require intervention. Allowing lower MAP during a free flap has very likely negative impacts on flap perfusion which is highly sensitive to perfusion pressure given that local innervation is erradicated when the flap is removed from its native neurovascular bundle. On the other hand, free flap surgeons often consider all vasopressors to be contraindicated.  This concept is not supported by general hemodyanamic physioglogy however. In particular, in patients whose stressed volume has been shifted to the unstressed vasculature after induction of and during maintenance of anesthesia, small doses of phenylephrine has been shown to increase cardiac output by shifting blood from the unstressed compartment (i.e. venous reservoir) into the stressed compartment. To state in other words, in normovolemic patients whose cardiac output is depressed due to decreased venous return, phenylephrine will predominantly increase blood pressure by increasing cardiac output via increased preload.  This increase in cardiac output will improve perfusion pressure to organs who are flow dependent for perfusion of oxygen. It has been documented, that in normal physiology, venous vasculature has a larger number of alpha receptors than that located on arteries and arterioles. Obviously, improving venous return via alpha 1 agonism depends exquisitely on ensuring normovolemia in your patients. Nevertheless, basic clinical research suggests that in general anesthesiologists should use vasopressors as needed in this patient population based on a perusal of a number of studies. For example Motakef et al. [3] published a systematic review of literature where they found that there is a high level of evidence that fluid should be between 3.5 ml/kg/hr up to max of 6 mL/kg/hr. Vasopressor use do not harm outcomes and may improve flap flow (highest level of evidence).  Chen et al. [4] performed a retrospective review on 187 patients and found that vasopressor use not associated with any negative consequences of any type. Eltorai et al. [5] also found this.  In a retrospective review they showed that Ephedrine treatment for hypotension during DIEP flap cases is associated with decreased intraoperative flap complication rates compared with controls who did not receive vasopressors, whereas phenylephrine has no significant association.  In a systematic review of studies done in head and neck free flaps the authors [6] looked at the effect of using vasopressors for BP control to determine if this increased risk of complications.   The review included one prospective and nine retrospective studies. Authors concluded that administration of phenylephrine and/or ephedrine was not associated with any complications which included: free flap failures, pedicle thrombosis, or other flap complications.  In a prospective observational study in 169 ENT free flap surgeries Monroe et al.  concluded that vasopressor use does not increase the risk of pedicle loss or complications in this population [7].   In yet another review of 320 patients who received vasopressors (phenylephrine and ephedrine) during free flap surgery for head and neck Harris et al. [8] determined that the administration of vasopressors were not associated with any complications or increased failure rate of flaps. In a different surgical population a retrospective review of 110 jejunal flaps for pharyngeal tumors was completed. Chen et al. [9]  were able to show that when phenylephrine and ephedrine were used there was no increase risk of complications or flap failures.

Therefore, optimizing perfusion of free flaps shares the same physiological approach utilized everyday by anesthesiologists whose is to optimize perfusion of all organs.  While it is true that due to denervation of the free flap perfusion pressure may not be automatically controlled by intrinsic bodily physiology, anesthesia providers can easily overcome this obstacle as they do in other scenarios (i.e perfusion of placenta in obstetric anesthesia).  Evidence suggest strongly that maintaining normovolemia by judicious use of a balanced crystalloid solution with careful titration of vasopressors when needed is the best approach to free flap microsurgery anesthesia based on current evidence.

1.  Karamanos E, Walker R, Wang HT, Shah AR. Perioperative Fluid Resuscitation in Free Flap Breast Reconstruction: When Is Enough Enough? Plast Reconstr Surg Glob Open. 2020 Mar 28;8(3)

2. Sjöberg, Thomas MD*,†; Numan, Anmar MD†,‡; de Weerd, Louis MD, PhD*,†. Liberal versus Modified Intraoperative Fluid Management in Abdominal-flap Breast Reconstructions. A Clinical Study. Plastic and Reconstructive Surgery - Global Open 9(9):p e3830, September 2021. | DOI: 10.1097/GOX.0000000000003830

3. Motakef S, Mountziaris PM, Ismail IK, Agag RL, Patel A. Emerging paradigms in perioperative management for microsurgical free tissue transfer: review of the literature and evidence-based guidelines. Plast Reconstr Surg. 2015 Jan;135(1):290-299. doi: 10.1097/PRS.0000000000000839. PMID: 25539313.

4. Chen C, Nguyen MD, Bar-Meir E, Hess PA, Lin S, Tobias AM, Upton J 3rd, Lee BT. Effects of vasopressor administration on the outcomes of microsurgical breast reconstruction. Ann Plast Surg. 2010 Jul;65(1):28-31. doi: 10.1097/SAP.0b013e3181bda312. PMID: 20548236.

5. Szabo Eltorai A, Huang CC, Lu JT, Ogura A, Caterson SA, Orgill DP. Selective Intraoperative Vasopressor Use Is Not Associated with Increased Risk of DIEP Flap Complications. Plast Reconstr Surg. 2017 Jul;140(1):70e-77e. doi: 10.1097/PRS.0000000000003444. PMID: 28654605.

6. Naik AN, Freeman T, Li MM, Marshall S, Tamaki A, Ozer E, Agrawal A, Kang SY, Old MO, Seim NB. The Use of Vasopressor Agents in Free Tissue Transfer for Head and Neck Reconstruction: Current Trends and Review of the Literature. Front Pharmacol. 2020 Aug 28;11:1248. doi: 10.3389/fphar.2020.01248. PMID: 32982724; PMCID: PMC7485519.

7. Monroe, M.M., Cannady, S.B., Ghanem, T.A., Swide, C.E. and Wax, M.K., 2011. Safety of vasopressor use in head and neck microvascular reconstruction: a prospective observational study. Otolaryngology--Head and Neck Surgery, 144(6), pp.877-882.

8. Harris, L., Goldstein, D., Hofer, S. and Gilbert, R., 2012. Impact of vasopressors on outcomes in head and neck free tissue transfer. Microsurgery, 32(1), pp.15-19.

9. Chan, J.Y.W., Chow, V.L.Y. and Liu, L.H.L., 2013. Safety of intra‐operative vasopressor in free jejunal flap reconstruction. Microsurgery, 33(5), pp.358-361. 10. Swanson EW, Cheng HT, Susarla SM, Yalanis GC, Lough DM, Johnson O 3rd, Tufaro AP, Manson PN, Sacks JM. Intraoperative Use of Vasopressors Is Safe in Head and Neck Free Tissue Transfer. J Reconstr Microsurg. 2016 Feb;32(2):87-93. doi: 10.1055/s-0035-1563381. Epub 2015 Sep 4. PMID: 26340760. 11. Fang L, Liu J, Yu C, Hanasono MM, Zheng G, Yu P. Intraoperative Use of Vasopressors Does Not Increase the Risk of Free Flap Compromise and Failure in Cancer Patients. Ann Surg. 2018 Aug;268(2):379-384. doi: 12.1097/SLA.0000000000002295. PMID: 28489683. 13. Milford EM, Reade MC. Resuscitation Fluid Choices to Preserve the Endothelial Glycocalyx. Crit Care. 2019 Mar 9;23(1):77. doi: 10.1186/s13054-019-2369-x. PMID: 30850020; PMCID: PMC6408751.