THE EFFECT OF BLOOD AROUND A FLAP PEDICLE ON FLAP PERFUSION IN AN EXPERIMENTAL RODENT MODEL

INTRODUCTION

Vasospasm is a commonly encountered physiologic response in flap and microsurgery which leads to impaired blood flow though the tissue. This phenomenon has been studied to a considerable extent in the neurosurgical literature, where it is one of the leading causes of morbidity and mortality following aneurysmal subarachnoid hemorrhage (1). However, little experimental and basic scientific literature about this potentially devastating response is available in the field of microsurgery. Flap failure has been attributed to a myriad of etiologies of which vasospasm and resulting thrombosis is well known, especially in cases where surgical technique, pedicle torsion, tension, kinking and/or external compression have been excluded (9). Vasospasm can affect the main pedicle vessels or the smaller vessels within the flap. In most cases, it resolves spontaneously, and no serious adverse effects result. However, if the response is severe and/or prolonged, ischemic necrosis can ensue, especially in poorly perfused tissue or distal aspects of a flap (19). The literature quotes a 6–25 per cent incidence of the above-mentioned complications (7, 11–14). In our clinical practice we have encountered a few cases of prolonged vasospasm which led to flap failures, and this prompted us to investigate this phenomenon in more detail.

In 1987 the senior author on the paper (JV) divided potential vasospasm of a flap and its pedicle into 4 categories: segmental vasospasm of the pedicle, diffuse vasospasm of the pedicle, vasospasm of the small vessels in the flap, and tissue shock of the flap and pedicle (15). Unfortunately, the physiology of vasospasm has still not been fully understood or studied. It is very likely that vasospasm at some level may be attributed to the loss of unsalvageable flap failures. We therefore designed an experiment in a rodent model to study vasospasm and its etiology during flap harvest. We were particularly interested in studying the effect of blood around the flap pedicle on flap perfusion. 

MATERIAL AND METHODS

We used forty adult male Wistar rats for our animal model. The mean weight of the rats was 300.4 grams. We followed the Greek and European Community guidelines for animal research as our standard. All surgery was performed under sterile, standard conditions. The animals’ core temperature was controlled at 25°C. They were anesthetized with a mixture of ketamine (100 mg/kg) and xylazine (10 mg/kg). After appropriate setup and preparation of the right groin, a pedicled groin flap based on the superficial epigastric artery and vein was harvested using only sharp dissection with scissors. Strict hemostasis was ensured by first clamping the edges of the proposed flap with hemostats. Care was taken to avoid tension on the pedicle, and all adventitia surrounding the pedicle was left undamaged. We measured a point 2 mm distal from the origin of the superficial epigastric vessels from the femoral vessels and cleared a 5 mm long segment of adventitia from the pedicle (Figure 1). This was done to prevent unequal thickness of adventitia in each animal and to standardize the exposure of the vessels to the stimulus. Next a laser-Doppler probe (PeriFlux system 5000, small straight probe 407-1, Perimed, Jarfalla, Sweden) was attached to the flap to measure continuous blood flow. At the same time the pedicle vessels were directly observed through an operating microscope for the presence of vasospasm during the experiment.

The rats were randomly divided into two groups [group A (n=20) and group B (n=20)] depending on the source of the blood to be used to bathe the flap pedicle in the experiment. Group A had blood originating from the tail of the rat by cutting the tip off and dripping it on the chosen segment of pedicle. Group B had blood originating from a bleeding branch of the femoral vessels. The total amount of blood used for the above purpose was negligible in terms of the blood circulation and hemodynamic stability.

After harvesting the flap and noting that the flowmetry readings were stable for at least 5 minutes (stable flap perfusion), we began recording the perfusion signals. The time point when the pedicle was exposed to the blood was termed “t=0”. The perfusion signal was recorded for 40 minutes, beginning from the time “t= –5” (5 minutes prior to exposure to the blood). An example of the recordings collected is shown in Figure 2. The perfusion recording signals were exported from the control software package of the laser-Doppler flowmeter into ASCII format files. Graphic representations of blood flow were created from these files (Figure 3). As the signals were corrupted by impulse noise it was impossible to detect properly the important time points and signal amplitudes. Therefore we employed a Savitzky-Golay polynomial filter to smooth the signals. We then extracted two important time periods “t_B” and “t_C” and two important signal amplitudes “v₁“and “v₂“ from the signals with use of Matlab scripts (see Figure 3). The time period “t_B” represented the period between “t=0” and the time point on the graph when perfusion began to increase after a period of poor perfusion secondary to the exposure of the pedicle to blood (see Figure 3). The time period “t_C” represented the period between “t=0” and the time when the re-perfusion reached its maximum level (see Figure 3). The signal amplitude “v₁“ was the perfusion unit value at the time just prior to “t=0”, and the signal amplitude “v₂“ was the lowest perfusion unit value after exposure of the pedicle to the blood (see Figure 3).

For each study animal all the mentioned signal characteristics – “t_B”, “t_C”, “v₁“ and “v₂” – were collected and compiled for analysis as shown in Figure 4 and Figure 5. Appropriate statistical analysis using the Wilcoxon Rank-Sum Test was preformed. A p-value of <0.05 was considered significant. 

RESULTS

Group A

The presence of blood (from the tail) around the pedicle caused a significant drop in perfusion (p=0.000019) compared to perfusion prior to the stimulus. The median decline in perfusion (v₂ – v₁) was –53.9 units. The median time period “t_B” was 525 seconds, and the median time period “t_C” was 1350 seconds.

Group B

The presence of blood (from a branch close to the pedicle) around the pedicle caused a significant drop in perfusion (p=0.000034) compared to perfusion prior to the stimulus. The median decline in perfusion (v₂ – v₁) was –71.3 units. The median time period “t_B” was 409 seconds and the median time period “t_C” was 948 seconds.

Comparison of Group A and B

In both groups there was a statistically significant drop in flow through the flaps when the pedicle was exposed to blood (p=0.0000000016) (see Figure 5). However, there was no statistically significant difference in the time periods “t_B” (p=0.33) (Figure 4a) and “t_C” (p=0.22)(Figure 4b) when the groups were compared.

DISCUSSION

Microsurgeons are all too familiar with the vasoconstriction associated with flap pedicle dissection. The pathophysiology of this reversible vascular response is still not fully understood. Vasospasm has been shown to cause structural changes and biochemical alterations at the level of the vascular endothelium, as well as surrounding smooth muscle cells and adventitia. Clinical evidence has suggested a number of potential etiologies for the vasospasm. Surgical manipulation of the pedicle itself, decreased temperature, and systemic circulatory changes have been mentioned most frequently (19). In the neurosurgery literature, blood in the subarachnoid space is believed to trigger vasospasm. Therefore one could assume that blood around  a flap pedicle could provoke a similar response during flap harvest. Thinking along these lines, some authors have investigated the effects of hematomas on healing of interpositional vein grafts in a rat model. These studies concluded that perivascular hematomas caused spasm and flow disturbances, prolonged vessel wall ischemia, and resulted in severe graft vessel wall injury. These negative effects delayed the healing process and subjected the grafts to the development of occlusive mural thrombi (2). To our knowledge, there has been no published data studying the effect of blood on pedicle vasospasm or flow.

Several experimental models have been described for studying vasospasm. Some models measure changes in the diameter of the vessel using in-vivo microscopy (3, 8, 16–18). Blood flow in a larger vessel has been measured more accurately using a Doppler probe (5). However, these models did not study changes in the microcirculation of the entire flap, as opposed to the more recent experimental models studied, which did (6). Our experimental model was designed to measure peripheral blood perfusion in the entire pedicled flap using both laser-Doppler flowmetry and direct observation of the vessels under a microscope. The potential stimulus for vasospasm (blood) was exposed only to the vessels (artery and vein) nourishing the flap. Therefore this design model more accurately reflects what is seen clinically and takes into consideration the possible changes in the periphery of the flap.

Experimental and clinical studies have shown that acute vasospasm is related to an acute ischemic response. The potential inducers of vasospasm are released from activated platelets as well as from mechanically damaged erythrocytes. Activated platelets release strong vasoconstrictors such as thromboxane A2 (TXA2), serotonin, ATP and platelet derived growth factor (PDGF). Some of them are known to stimulate the release of a powerful vasodilator nitric oxide (NO) from endothelial cells. NO relaxes vascular smooth muscle cells directly and indirectly by down-regulation of the production of two powerful vasoconstrictors – endothelin-1 (ET-1) and 20-hydroxyeicosatetraenoic acid (20-HETE) – from the vascular wall. Black et al. have shown that the contractile response to ET-1 is significantly higher in human veins than arteries, thus supporting the clinical impression that veins are more susceptible to vasospasm than arteries in the vascular pedicle of flaps (18). Theoretically, this NO dependent vasodilatory effect can be mitigated if NO is scavenged by binding to hemoglobin, thus decreasing its availability. “Scavenging hemoglobin” is made available when there is fresh blood or clot and mechanically damaged RBCs around the flap pedicle. This is one theoretical mechanism by which blood around the flap pedicle could lead to decreased flap blood flow. Another theory which attempts to account for vasospasm is based on an increase in the concentration of endothelin-1 in plasma during acute ischemia. This may result from the direct stimulation of its production by the endothelial cells by oxyhemoglobin or vasopressin. Thus there may be many different mechanisms that result in acute vasospasm/ischemia; however, a key point seems to be the impairment of nitric oxide-dependent regulation (10).

This experiment verifies that blood can cause significant vasoconstriction of the pedicle and decrease flow through a flap. In clinical terms, it supports the hypothesis that strict hemostasis around a flap pedicle and prevention of hematomas is essential during surgery to mitigate this potentially devastating effect. This is especially important in perforator flaps and digital replants/transfers, where the diameter of the pedicle and its branches are extremely small and flap loss is potentially higher with severe/prolonged vasoconstriction. In order to better understand the mechanisms which play a role in development of a vasospasm, further investigation of the effect of vasoconstrictor and vasodilator substances is needed. 

CONCLUSIONS

Our study clearly shows that the presence of blood around the pedicle of a flap can produce significant deterioration in perfusion through vasospasm. It also indicates that the source of bleeding is not of significant importance. Based on these findings, we recommend that every effort should be made to perform dissection of the flap pedicle in a bloodless field and to minimize post-operative accumulation of blood around the flap pedicle. Similarly, in the event of flap ischemic issues, vasospasm of the pedicle vessels should be anticipated in addition to ruling out kinking, torsion, tension and external compression of the pedicle.

Acknowledgment:

This work was supported by the grant IGA NR-8368-5.

Address for correspondence:

Petr Hýža, M.D.

Berkova 34

612 00 Brno

Czech Republic

E-mail: petr.hyza@fnusa.cz