Introduction
The problem of anastomotic thrombosis in microsurgical flaps occurred immediately after the world’s first microsurgical transplantation of free deepidermized epigastric cutaneous-fatty flap for closure of the right-sided lateral facial defect in a 35-year-old woman. Flap reperfusion was achieved through the branch of external carotid artery (end-to-end) and internal jugular vein (end-to-side). Surgery was complicated by venous thrombosis [1].
In 1996, the complications following postoperative anastomotic thrombosis were analyzed for the first time in a large clinical material (transplantation of 990 free flaps). The incidence of anastomotic thrombosis was 5.1%. Vein thrombosis was more common compared to arterial thrombosis (54% vs. 20%, thrombosis of both vessels (veins and arteries) — 12%) [2]. According to meta-analysis data (PubMed and Embase databases for the period 2000 — 2018; 283 flaps), the incidence of postoperative complications (partial or total necrosis of a free flap) was 6%. Of these, 5% of complications (excluding wound suppuration) occurred on the background of venous and arterial anastomotic thrombosis. It is noteworthy that the incidence of marginal or total necrosis after free flap transplantation on the upper extremities was significantly higher compared to transplantation on the lower extremities (8% vs. 6%). Moreover, 95% of all flap necrosis were associated with “anastomotic” thrombosis of veins or arteries [3]. The highest incidence of vascular complications was observed in patients with head and neck tumors. Free transplantation of tissue complexes was followed by arterial and venous thrombosis in 11.7% of cases, necrosis — in 17.6% of cases [4]. In 2019, the authors confirmed the feasibility of 2 venous anastomoses for flap drainage after reperfusion in repair of lower limb defects. Incidence of complications (partial or total necrosis) in drainage of anterolateral thigh (ALT) flaps through one vein was significantly higher compared to drainage through two veins (47% vs. 24%; p=0.065) [5].
Currently, microsurgeons have focused their attention mainly on technical aspects of microvascular anastomoses. In their opinion, improvement of surgical technique should reduce the incidence of postoperative complications (congruence of vascular ends (mechanical sutures) and clarifying the indications for two types of manual sutures (“end-to-end” or “end-to-side”) [5-8].
At the dawn of microvascular surgery, some experimental studies were devoted to the processes of arterial and venous wall regeneration within patent microvascular anastomoses in “end-to-end” [9, 10] and “end-to-side” fashion [10].
Meanwhile, we found no reports devoted to the causes of venous thrombosis from the standpoint of pathological anatomy and physiology of hemodynamic disorders in free flaps.
The purpose of this research was experimental study of pathological anatomy and physiology of venous system in the model of transplantation of non-free and free fasciocutaneous flaps.
The objectives of the study:
- experimental research of venous outflow on the model of epigastric flap;
— creation of geometric shape of axial (epigastric) vein and its venules;
— animation of venous wall mobility before and after microvascular suture;
— modeling of erythrocyte flow in veins before and after microvascular suture.
Material and methods
There were 5 series of experimental studies (125 white rats, 25 animals in each sample). We analyzed the following aspects:
— venous outflow in a locally displaced non-free epigastric flap with intact vascular pedicle (group 1) (Fig. 1);
— venous outflow in a locally displaced non-free epigastric flap after desympathization of the axial artery wall (group 2);
— venous outflow in a locally displaced non-free epigastric flap after selective desympathization of the axial vein (group 3);
— venous outflow in a locally displaced non-free epigastric flap after combined desympathization of the entire vascular pedicle of the flap (group 4);
— venous outflow in a free epigastric flap reperfused by the method of P. van der Sloot (2002) (group 5) (Fig. 2).
Fig. 1. Stages of non-free epigastric flap displacement.
a — flap elevation (stage I); b — medial reversal of the flap by 90° (stage II).
Fig. 2. Modeling of free epigastric flap by P. van der Sloot (2002).
a — view before transplantation; b — view after microsurgical stage.
Skin of epigastric region of 10 white rats was used as a control material.
The choice of epigastric fasciocutaneous flap for experimental studies was caused by its advantages: thinness; one valveless axial vein in vascular pedicle; subdermal location in the same plane of the branches of axial artery and axial vein; abundance of arteriovenular anastomoses in the skin.
We used standard epigastric flaps (2×3 cm) in white rats of both sexes to study the response of venous system of fasciocutaneous flap to various influences on the vascular pedicle. At different times after graft transplantation (3, 7, 14 days), retrograde filling of veins with blue Gerota mass was performed through the femoral vein under 60–70 mm Hg. Then, the flaps with stained veins were clarified using the method of V. Shpaltegolts (1921) modified by D.A. Zhdanov (1943). Preparation of microanatomical and histological specimens was performed at the final stage.
We harvested skin fragments in 3 points (central, peripheral and at the entry of the vascular pedicle into the flap) at different times after graft transplantation to study tissue reactions in the skin of the flaps. Specimens were fixed in a 12% solution of neutral formalin and Carnoy’s fluid and embedded in paraffin. Dewaxed histological specimens were stained with haematoxylin and eosin. To obtain semi-thin slices, we fixed specimens in 2.5% glutaraldehyde buffered in 0.2 M cacodylate buffer (pH 7.2) with post-fixation in 1% osmium tetroxide solution and embedded these specimens in araldite. Slices (LKB-III ultratome) were stained with toluidine blue.
Statistical analysis was performed using Statistica 7.0 for Windows and SPSS Statistics 17.0 software package (Spearman’s correlation coefficient). Qualitative variables were compared by using of Fisher’s exact test.
Modeling, animation and rendering were done Blender 2.79 software package. Geometric shape of the vessel was created using a simplified low-poly model with subsequent edge sharpness input. The entire model was smoothed using Catmull-Clark method (Subsurf modifier). Creation of venules was based on a simplified geometry with edges only (no planes). After that, these models were processed using Skin modifier. Animation of vascular wall mobility (muscle contractions) was achieved using Wave geometric modifier. Denervated vascular wall movements in the diastolic phase (before microvascular suture) were animated using Displace geometric modifier and dynamic displacement map. Erythrocyte emitter transformation (decrease — increase) was obtained synchronously with venous wall movement. Erythrocyte flow was modeled using ParticleSystem-Fluid program incorporated into Blender software package.
Mathematical formulation of the problem. Overall mass balance for each phase (k = 1 for plasma and k = 2 for erythrocytes) is described by the equations:
where ρ — density, εk — volume fraction of the kth fraction, t — time, 
— speed. Moreover, the sum of volume fractions of the fractions of each phase should be equal to one:
,
where np — total number of phases. The volume of one fraction cannot be occupied by another fraction.
The momentum equation for each of the phases will be written as:
,
where p — pressure, 
— viscous stress tensor, 
— acceleration of gravity, 
— vector of external forces. In the force of interphase interaction, k and l represent the coefficients corresponding to plasma and red blood cells, βki — coefficient of interaction of phase motion moments.
Rheological model of blood. Rheological model of blood is the main objective in hemodynamic modeling. The non-Newtonian stress model is able to describe the flow of dense suspension with a large amount of inclusions. In our model, dimensionless relative viscosity of the mixture η depends on the shear stress and hematocrit, as it was suggested in a Ding model. Dimensionless relative viscosity of the mixture can be calculated from the experimental rheological data [11]:
,
where m, λ and n — model constants defined below.
Strain rate: 
.
Dimensionless value of strain rate:
.
Experimental results were interpreted considering morphofunctional features of axial vessels of epigastric flaps in rats:
1) vascular pedicle of epigastric flap includes an artery and only one vein located in a common vascular vagina;
2) axial vein of epigastric flap belongs to muscle type veins with a weak development of muscle elements in the middle layer of the wall;
3) axial vein wall in vascular pedicle, in contrast to the artery, is supplied only by vasa vasorum externa arising from adjacent axial artery;
4) there are no valves in the axial vein of epigastric flap;
5) perivenous nerve plexus of axial vein is in close anatomical relationship with vasa vasorum externum;
6) epigastric artery pulsation (axial artery of muscular type) is involved in motor activity of adjacent axial epigastric vein.
Results
There was a well-defined venous bed on skin specimens within the area of epigastric flap dissection. Veins forming an axial vein of the flap had a slightly tortuous course with an outer diameter of 67.7 [57.4; 70.1] μm (Fig. 3).
Fig. 3. Venous system of skin in epigastric flap.
Injection of blue Gerota mass. Magn. ×16.
Group 1 (intact vascular pedicle). Retrograde injection of dye in 3 days after surgery revealed a random arrangement of peripheral veins of the flap. In the central part of the flap, veins were ordered, enlarged and convoluted; dermal microvasculature was not stained (Fig. 4). By the 5th day after surgery, we observed a network of different-sized anastomoses between venous vessels. The last ones were characterized by significant tortuosity.
Fig. 4. Venous system of central part of displaced non-free epigastric flap (with untreated vascular pedicle), 3 days after surgery.
Enlarged and convoluted axial vein in the center of the flap. Injection of blue Gerota mass. Magn. ×16.
On the 7th day, central veins of the flap were enlarged, microvasculature was stained. By the 14th day, central veins of the flap forming an axial vein of vascular pedicle were rectilinear, microvasculature was well stained (Fig. 5).
Fig. 5. Venous system of central part of displaced non-free epigastric flap (with untreated vascular pedicle), 14 days after surgery.
Dilated major tributaries of the axial vein in the center of the flap. Injection of blue Gerota mass. Magn. ×16.
Morphometry of the intra-flap veins of locally displaced non-free epigastric flaps with intact vascular pedicle revealed reduction of their diameter within 3 days after surgery and subsequent significant enlargement (p<0.05). Tendency to vein enlargement persisted up to the 14th day after surgery (Table).
Table. Diameter of axial arteries and veins (μm) of epigastric flap after its transposition with an intact vascular pedicle, Me [Q25; Q75]
|
Day |
Group 1 |
Group 2 |
Group 3 |
Group 4 |
Group 5 |
|
3 days |
45,3 [34,2; 49,5]#* |
64,0 [59,1; 67,4]#† |
41,2 [34,3; 50,6]#†* |
35,1 [31,2; 36,0]* |
39,0 [35,8; 40,1]#†* |
|
7 days |
83,1 [72,2; 84,0]^* |
89,1 [81,4; 90,3]^†* |
85,0 [79,0; 86,3]^* |
48,1 [34,4; 66,1] |
89,0 [84,6; 90,5]^†* |
|
14 days |
71,0 [68,0; 72,1]^ |
74,2 [71,1; 80,3]^# |
76,0 [74,1; 82,0]^* |
41,3 [34,0; 43,5]* |
71,2 [67,3; 72,8]^# |
|
Control |
67,7 [57,4; 70,1]^ |
67,7 [57,4; 70,1]^ |
67,7 [57,4; 70,1]^ |
67,7 [57,4; 70,1]^ |
67,7 [57,4; 70,1]^ |
Note. Significant differences (p <0.05): ^ — compared to the 3rd day; # — compared to the 7th day; † — compared to the 14th day; * — compared to the control group.
In 3 days after surgery, skin of the peripheral part of the flap was characterized by a weak expression of papillary layer and chaotic arrangement of collagen fibers of different thicknesses. There was a small amount of elastic fibers. Cellular composition included fibroblasts and small number of poorly differentiated cells. At the early stages (up to 7 days), skin of the flap had significant lymphocytic-histiocytic infiltration. By the 14th day, edema, venous congestion and enlarged veins persisted in the dermis of the flap. Blood cells and erythrocyte sludge complexes were visualized in some veins (Fig. 6).
Fig. 6. Derma of non-free epigastric flap (with untreated vascular pedicle), 14 days after surgery.
Sludge complexes are visible in the lumen of the veins. Toluidine blue staining. Magn. ×900.
Group 2 (axial artery desympathization). Retrograde filling of veins of epigastric flap through an axial vein on the 3rd day after surgery revealed a great number of tortuous and dilated veins in the central zone of the flap. Staining of microvasculature was weak (Fig. 7). In 5 days after surgery, staining of microvasculature remained weak. The branches of the axial vein were moderately dilated.
Fig. 7. Venous system of skin of a non-free epigastric flap after periarterial sympathectomy of the vascular pedicle, 3 days after surgery.
Convoluted and dilated axial vessels are clearly visualized in the central part of the flap. Injection of blue Gerota mass. Magn. ×16.
On the 7th day, anastomoses between large venous branches were well visualized in the central part of the flap. Active staining of microvasculature was observed over the entire skin of the flap. By the 14th day, branches of the axial vein remained dilated and extremely tortuous in the skin of the central part of displaced non-free epigastric flap with a well-stained microvasculature (Fig. 8).
Fig. 8. Venous system of skin of central part of non-free epigastric flap after periarterial sympathectomy of the vascular pedicle, 14 days after surgery.
Convoluted and dilated branches of the axial vein in the center of the flap. Injection of blue Gerota mass. Magn. ×16.
The diameter of the main branches of the axial vein of the epigastric flap after periarterial sympathectomy of the axial artery was characterized by initial mild dilation (3 days after surgery) and further significant enlargement (p <0.05) up to the 14th day (Table).
On the 3rd day, partial violation of layer architectonics was observed in the skin (poorly differentiated papillary layer and few elastic fibers). These signs persisted until the 14th day. There were severe tissue edema and venous plethora in the flap. Blood cells and sludge complexes were visualized in some dermal veins (Fig. 9).
Fig. 9. Vein of reticular dermis of non-free inguinal flap with periarterial sympathectomy of the vascular pedicle, 14 days after surgery.
Sludge complexes of red blood cells (arrow). Toluidine blue staining. Magn. ×900.
Group 3 (axial vein desympathization). On the 3rd day after displacement, the main branches forming an axial vein were enlarged and convoluted in the central part of non-free epigastric flap (Fig. 10), microvasculature was not stained. In 7 days after surgery, microvasculature was not stained after retrograde filling of venous bed. Venous bed of the flap was represented by dilated vessels of various dimensions.
Fig. 10. Venous system of skin of central part of non-free epigastric flap after perivenous sympathectomy of v. epigastrica superficialis, 3 days after surgery.
Convoluted and dilated veins in the center of the flap. Injection of blue Gerota mass. Magn. ×16.
In 14 days, dilatation of the main tributaries of the axial vein persisted in the skin of the flaps (Fig. 11). Perivenous sympathectomy was followed by early significant enlargement of the main tributaries of the axial vein (p<0.05), i.e. since the 3rd day after surgery. Vasodilation persisted up to 14 days (Table).
Fig. 11. Venous system of skin of central part of non-free epigastric flap after perivenous sympathectomy of v. epigastrica superficialis, 14 days after surgery.
Dilated branches of the axial vessels are visualized in the center of the flap. Injection of blue Gerota mass. Magn. ×16.
In the skin of the flap, we observed disorientation of collagen fibers, smoothness of papillary layer and few elastic fibers in early postoperative period (up to 5 days). There were local accumulations of histiocytes and leukocytes in the papillary layer, as well as edema and venous congestion with enlargement of most venous and lymphatic vessels. Blood cells and erythrocyte sludge complexes were visible in veins (Fig. 12).
Fig. 12. Papillary dermis of non-free epigastric flap after perivenous sympathectomy of v. epigastrica superficialis, 7 days after surgery.
Phenomenon of venous congestion: sludge complexes in the lumen of the vein. Toluidine blue staining. Magn. ×16.
Group 4 (simultaneous desympathization of axial artery and vein). In 3 postoperative days, we visualized dilatation and tortuosity of the main branches of the axial vein in the central part of the flap; staining of microvasculature was weak. Various-sized veins were located in the peripheral part of the flap (Fig. 13).
Fig. 13. Venous system of skin of central part of non-free epigastric flap after sympathectomy of a. et v. epigastrica superficialis, 3 days after surgery.
Enlarged branches of the axial veins. Injection of blue Gerota mass. Magn. ×16.
After 7 and 14 postoperative days, microvasculature was well stained over the entire skin of the epigastric flap. The axial vein tributaries were dilated. In the peripheral part of the flap, we found well-stained venous bed with a system of intervenous anastomoses (Fig. 14).
Fig. 14. Venous system of skin of central part of non-free epigastric flap after sympathectomy of a. et v. epigastrica superficialis, 14 days after surgery.
Rectilinear course of the branches of axial vessels. Injection of blue Gerota mass. Magn. ×16.
Simultaneous sympathectomy of superficial epigastric artery and vein was followed by decrease in diameter of the main tributaries of the axial vein within 3 days after surgery and their subsequent significant enlargement (p <0.05). Vasodilation persisted up to 14 days (Table).
There was no clear dermal architectonics in the flaps after 3 postoperative days (unclear layers with chaotic arrangement of collagen fibers). Focal leukocyte-histiocytic infiltration was observed in deep layers of dermis. Edema and enlarged vessels with sludge complexes were found throughout the flap. Signs of edema and inflammation persisted throughout the entire postoperative period and regressed only by the 14th postoperative day (Fig. 15).
Fig. 15. Reticular layer of the dermis of non-free epigastric flap after sympathectomy of a. et v. epigastrica superficialis, 14 days after surgery.
Numerous blood cells are visible in the lumen of dilated vein. Toluidine blue staining. Magn. ×900.
Group 5 (free reperfused flap). In 3 days after surgery, retrograde filling of venous bed revealed predominance of enlarged veins of medium diameter (tortuous tributaries of the axial vein with a large number of intervascular anastomoses). Staining of microvasculature was weak (Fig. 16).
Fig. 16. Venous system of free epigastric flap, 3 days after surgery.
Sharply enlarged axial vein in the center of the flap. Injection of blue Gerota mass. Magn. ×16.
These vascular features (dilated and convoluted main tributaries of the axial vein, weak staining of microvasculature) persisted in the skin of the free reperfused epigastric flap until the 14th day after surgery (Fig. 17). Diameter of the branches of the axial vein was reduced within 3 days after surgery. Significant enlargement was observed later (p <0.05). Vasodilation persisted until the 14th day (Table).
Fig. 17. Venous system of free epigastric flap, 14 days after surgery.
Enlarged and convoluted axial vein in the center of the flap. Injection of blue Gerota mass. Magn. ×16.
In early postoperative period, we observed disorientation of dermal layers, smoothness of papillary layer and chaotic arrangement collagen fibers of different thickness. Elastic fibers were almost undetectable. Cellular dermal composition of the free flap included fibroblasts and poorly differentiated cells. In 3 days after surgery, we observed tissue edema and venous plethora in all areas of the flap; arterial and venous vessels were dilated. Tissue edema and venous congestion persisted until the 7th day after surgery. Intravascular erythrocyte sludge complexes and small number of capillaries were visualized in all parts of the flap (Fig. 18).
Fig. 18. Free epigastric flap, 7 days after surgery.
Blood cells (sludge complexes) in the lumen of the venule. Hematoxylin and eosin staining. Magn. ×300.
Thus, experimental studies have shown a high sensitivity of venous bed of epigastric flap to any surgical manipulation including usual dissection (group with intact vascular pedicle). Surgical manipulations with vascular pedicle of epigastric flap (various options of axial vessel desympathization) and its intersection with subsequent reperfusion of the flap are accompanied by obvious venous and microvascular reactions.
Early venous response to selective desympathization of the axial artery was unexpected. This reaction was followed by progressive enlargement of the convoluted veins up to the 14th day. The same changes were characteristic for the microvasculature: from weak (3 days) to active (14 days) staining.
Reaction of venous bed of the flap on selective desympathization of the axial vein was is specific (early and significant dilatation of tortuous veins persisted until the 14th postoperative day). There was no staining of microvasculature until the 3rd day. We observed weak staining only by the 7th day with its subsequent increase by the 14th day.
Venous reaction of the flap to simultaneous desympathization of the axial artery and vein is not a mechanical summation of separate vascular reactions after selective desympathization of the artery and vein. This reaction included 2 phase (initial reduction of the tributaries of the axial veins within 3 days and subsequent significant enlargement up to the 14th day). Staining of microvasculature was changed from mild (the 3rd day) to bright coloring after 7–14 days.
Venous response of a free (denervated) epigastric flap after reperfusion of a free flap differs from venous response in a non-free epigastric flap with desympathization of both vessels in vascular pedicle. In a free flap (primary ischemia, 60–90 min), reaction was characterized by early dilatation of convoluted veins with a large number of intervascular anastomoses. Staining of microvasculature was weak throughout the entire follow-up period.
Interpretation of our data on veins and microvasculature in the epigastric flap was based on assumption that enlargement and tortuosity of the tributaries of valveless axial vein indicate an increase in venous bed capacity and blood flow slowdown. Staining of microvasculature characterizes venous hypertension in the system of valveless axial vein. Histological data confirm this interpretation, since tissue edema and erythrocyte sludge are the main indicators of impaired peripheral hemodynamics in tissues due to venous stasis (hypertension) and impaired laminar blood flow in denervated venous bed.
Considering these result, as well as previous data on arterial perfusion of free flaps, we performed computed animation of arterial inflow and venous outflow from the reperfused axial fasciocutaneous (epigastric) flap using mathematical modeling of cardiovascular biomechanics (Fig. 19). Arterial inflow through the axial artery (after microvascular suture) occurs in the absence of cardiac synchronization of its motor activity. This is accompanied by impaired neurogenic mechanism of opening of arterioles regulating capillary perfusion of the flap. Another aspect is impairment of rheological properties of blood. In this regard, laminar blood flow through the axial artery is disturbed within and distal to vascular anastomosis. Decrease in erythrocyte flow rate us also observed (normally 0.5–1.0 mm/s). Venous outflow from the reperfused flap is accompanied by dilatation of the tributaries of the axial vein, venous congestion, decrease in venous blood flow velocity and erythrocyte sludge complexes in venules and tributaries of the axial vein even in 7–14 days after surgery.
Fig. 19. Computer modeling and animation of venous blood flow infasciocutaneous flap after reperfusion.
The animation can be viewed using the QR code below the picture.
Discussion
In recent years, physiologists have paid a great attention to the role of cardiac output in peripheral vascular perfusion, as well as the role of peripheral vascular bed in circulatory regulation. These findings are very important for anesthesiologists. Indeed, the last ones have become increasingly aware of their role as clinical pharmacologists in improvement of the outcomes of free graft transplantation. Normally, overall vascular resistance in systemic circulation is distributed as follows: elastic vessels (great arteries) — 18%, muscular arteries (medium-sized arteries and arterioles) — 50%, capillaries — 25%, venules with a diameter > 50 μm — 4%, other veins — 3%. Peripheral vascular resistance in tissues is caused by friction of cells against the vascular wall and against each other (viscosity). Resistance is significantly increased in denervated vessels of reperfused flaps. There is still no effective anesthetic management to reduce peripheral vascular resistance in free flaps caused by concentration of denervated vessels. Most of these vessels in free flaps (medium-sized arteries and veins, arterioles and venules with a diameter > 50 μm) have lost neurogenic control of their resistive function. Only local endothelial and humoral mechanisms of regulation of resistive function are preserved. These violations of vascular motor activity lead to increase in peripheral vascular resistance and impairment of rheological properties of blood in the flap. Peripheral vascular resistance is acutely increased in both inflow and outflow vessels of the flap. In afferent vessels, it is caused by impairment of arteriole opening mechanism and reduced capillary perfusion of the flap, as well as violation of laminar arterial blood flow within and distal to vascular anastomosis. In outflow vessels (venules and veins), hemodynamic conditions differ significantly from those in arteries and arterioles. Collecting venules (>50 μm) are formed after confluence of postcapillary venules. These vessels normally have a resistive function underlying drainage of numerous non-muscular postcapillary venules. Normally, mean linear blood flow velocity in collecting venules is 6–7 times higher than in capillaries (venules — 0.3–1.0 cm/s, capillaries — 0.05-0.07 cm/s). Venous wall denervation decreases cross-section of the vessels, increases venous bed capacity by 20% and significantly reduces linear blood flow velocity in the veins [12–14].
This information is basic for discussion of restoration of adequate venous drainage in free flaps. Currently, free tissue complexes are often used to close extensive defects of the head and neck, upper and lower extremities, etc. The most relevant problems include: 1) state of venous bed in free autograft; 2) selection of recipient veins; 3) design of vascular anastomosis between the donor and recipient veins; 4) prevention of venous anastomotic thrombosis.
We obtained original experimental data on the state of venous bed in free (reperfused) epigastric flap. Indeed, we found predominant enlarged medium-sized veins with significant tortuosity, a large number of intervascular anastomoses and significant enlargement of venous vessels in the flap. Importantly, the factors predisposing to thrombosis are initially created in venous bed of a reperfused flap, in particular, interrupted blood flow during the period of primary ischemia of the flap. Laminar blood flow (erythrocyte sludge complexes) impairment, increase of peripheral vascular resistance and reduced linear blood flow velocity are observed in the denervated donor veins. All these features results slowdown of venous outflow from the flap and venous plethora after reperfusion [15, 16]. Rapid filling of venous bed after flap reperfusion can indicate a blockage of microcirculatory bed and intensive arteriovenous shunting.
The choice of recipient veins is technical and physiological aspect of surgery since the dosed “suction” function of recipient vein will be important for the denervated flap. Therefore, physiological features of venous outflow in recipient veins are very interesting for the specialists in reconstructive microsurgery.
Venous outflow from the head, neck and upper extremities is performed into superior vena cava through the jugular veins (head, neck), deep and superficial veins of the upper extremities. Cardiac cycles and respiratory phases affect venous return to the right heart [12, 17]. Venous pressure in internal jugular veins rapidly drops to zero in the first third of inspiratory phase, and then rises in the second half of inspiratory phase [17, 18].
Reconstructive surgeons dealing with head and neck tumors should pay attention to at least two things:
1) ideal venous anastomosis implies the use of two recipient veins for anastomosis with two comitant donor veins (“end-to-end”) or one large recipient vein whose diameter is 2 times greater than the diameter of the donor vein (“end-to-side”) [5, 7];
2) peculiar respiratory reaction of internal jugular vein as a possible recipient vein should be considered in surgery of head and neck tumors. Venous pressure in internal jugular veins quickly drops to zero in the first third of inspiratory phase, and rises in the second half of inspiratory phase.
Physiological feature of venous outflow from the lower extremities is high susceptibility of venous capacity to blood deposition under orthostatic load. In horizontal position, venous pressure in the lower extremities is only 10-15 mm Hg. In upright position, distal venous pressure in the lower extremities is increased up to 85-100 mm Hg depending on the patient’s height. This is true for both superficial and deep veins of the lower extremities. However, blood deposition is observed in deep veins (plus 300–400 ml) rather superficial ones due to anatomical features of the lower limb venous system (mainly muscular venous sinuses) [17]. Permanent tension of skeletal muscles of the lower extremities is necessary to maintain the posture. This process is accompanied by intramuscular pressure increase by 50–60 mm Hg. This is sufficient to limit vein enlargement and prevent orthostatic disorders. However, function of musculovenous pump is more important for venous return [19]. Respiratory movements (inspiration) and, to a lesser extent, suction forces arising from right atrium contraction, significantly affect venous outflow from the lower extremities [12, 17].
Reconstructive surgeons should consider the methods for improvement of deep venous drainage from the lower extremities and, accordingly, free flaps in this area:
1) intraoperative segmental intermittent pneumatic compression (SCD EXPRESS, Tyco Healthcare/Kendall, USA), the elevated position of the limb and breathing exercises in postoperative period;
2) effective venous drainage of free flaps, transplanted for closure of the lower limb defects, is ensured by 2 venous anastomoses. In these cases, the incidence of venous anastomotic thrombosis is minimal [5].
Design of vascular anastomosis between the donor and recipient veins has been actively discussed since 2001 [6, 7, 20, 21]. Venous anastomosis in end-to-side fashion is preferred for reconstructive microsurgery of lower limb defects. This conclusion was made considering the incidence of venous thrombosis for anastomoses in end-to-end and end-to-side fashion [5]. In 2019, the authors tried to objectify the choice of anastomosis design in patients with soft tissue defects after resection of head and neck malignancies (oral cancer, tongue cancer) [8]. The criterion was blood flow velocity in donor and recipient vein. Importantly, blood flow velocity may be different for the same flap, for example, ALT flaps in different patients. The researchers proceeded from the fact that a perfect vascular suture (i.e. absolutely congruent anastomosis) is impossible, especially in venous anastomoses. Vascular suture will necessarily deform thin venous wall within the anastomosis and impairs blood flow (occurrence of a recirculation zone (“swirls”) within the anastomosis). It was proven that the choice of venous anastomosis design depends not so much on the diameter of donor vessel or thickness of its wall, but on intravascular blood flow velocity. According to digital simulation data, blood flow recirculation within the end-to-end anastomosis was absent in case of low blood flow velocity (4.2 ml/min in the donor vein and 6.0 ml/min in the recipient vein). High blood flow velocity in the donor vein of the ALT flap (up to 24 ml/min) was followed by recirculation zone within end-to-end anastomosis with recipient local vein (superior thyroid vein). In this case, end-to-side anastomosis with internal jugular vein is preferable (blood flow velocity 400 ml/min). There were no recirculation zones within the anastomosis with jugular vein [8]. In the future, intraoperative ultrasound of donor and recipient veins may be used for prevention of anastomotic venous thrombosis. This procedure will be valuable to determine the type of microvascular anastomosis in each case.
Currently, technical aspects of microsurgical anastomoses are well developed. However, the incidence of postoperative complications after transplantation of free flaps (anastomotic thrombosis) is still high. As always, venous thrombosis prevails. It is believed that early postoperative antithrombotic therapy is mandatory in reconstructive microsurgery. Unfortunately, the hopes for effective prevention with aspirin did not come true [22]. Combination of anticoagulation with heparin and antiplatelet therapy should be applied individually [23]. The Consensus for the prevention of thrombosis in reconstructive microsurgery should be based on not only postoperative therapy, but also fundamental pathophysiological and pathomorphological data on arterial, venous and microcirculatory beds in free flaps. These features differ somewhat in the inflow and outflow vessels [15]. According to modern concepts, local blood flow regulation in denervated flap involves humoral substances released by endothelium in response to primary ischemia, as well as cellular elements (erythrocytes, platelets, neutrophils) [24]. Considering these data, we can understand the differences in blood clots (“white thrombus” of platelets and leukocytes in arteries and “red thrombus” of erythrocytes and fibrin in veins). Erythrocyte sludge in veins of the flaps is associated with changes in blood rheological properties (fluidity, viscosity) caused by increase in peripheral vascular resistance and laminar blood flow impairment in axial vessels.
In 2008, the Consensus on the Prevention of Venous Thromboembolism was adopted (“Virchow’s triad” in honor of the famous German pathologist of the 19th century). At first, Virchow’s triad included interrupted blood flow, injury of vascular wall and surrounding tissues, and hypercoagulability. The latest version of the triad is as follows: stasis of the blood flow, endothelial injury, hypercoagulability. The entire Virchow’s triad is presented in case of transplantation of free grafts with autonomous vascular blood flow. According to multichannel laser Doppler flowmetry, “steal phenomenon” was observed in skin of free flaps after epidural anesthesia. This phenomenon was followed by blood flow reduction in the flaps by 20–30% (p<0.05) without vascular reactions in intact skin and muscles around the flap [25].
Conclusion
The lost reperfused free flap is a great tragedy for a patient and disappointment of surgeon due to the lack of a result of surgical treatment. Moreover, this is an enormous financial cost for a hospital. Struggle for a viability of the transplanted free flap begins at the intraoperative stage and anesthesiologists are essential in these processes. Anesthetic measures are standard and include effective analgesia, maintaining a normal body temperature and adequate perfusion pressure, high cardiac output (systolic pressure above 100 mm Hg) and low systemic vascular resistance to prevent or reduce peripheral vasoconstriction in the flap, mild hemodilution (hematocrit 30–35%) for better microcirculation in the flap. Unfortunately, these procedures are carried out without considering the fact that the reperfused free flap is autonomous in local blood flow nervous regulation, insensitive, venous bed is paralyzed. Arterial bed responds to the pulse wave, but the lost neurogenic control of basal arterial tone distal to vascular suture (with preserved endothelial and myogenic one) is accompanied by impairment of opening of arterioles in response to left ventricular contraction. These processes reduce capillary perfusion in the flap. An increase of peripheral vascular resistance, venous plethora and tissue edema develop in the reperfused flap. Erythrocyte sludge complexes in venules and veins are observed even in 7–14 days after surgery. These circumstances require clarification of the generally accepted standard for intraoperative management of “microsurgical” patients. Regarding prevention of anastomotic venous thrombosis, we believe that prevention should imply fulfillment of the requirements of the Consensus on prevention of thrombosis, correction of blood rheological disorders and individual approach to microvascular venous anastomosis design (“end-to-end”, “end-to-side”) based on blood flow velocity.
The authors declare no conflicts of interest.