British _%urnal
ofPlastic
Surgery (1979),
JZ, 158-163
PATENCY OF ANASTOMOSES ADJACENT TO THE BIFURCATION OF THE RABBIT FEMORAL ARTERY AND A COMPARISON WITH Y-SHAPED MICROARTERIAL GRAlW3 By D. De La PAVA, M.D., GEMMANIGHTINGALE, M.Sc.(Melb), B. B. SHAFIROFF, M.D. and B. McC. O’BRIEN,B.Sc., M.D., MS., F.R.C.S., F.R.A.C.S., F.A.C.S. Microsurgery Research Unit, St Vincent’s Hospital, Melbourne, 3065, Australia
THE major cause of failure in replantation and microvascular free tissue transfer surgery in our unit has been arterial thrombosis and a number of causative factors have been identified. It has been suggested that arterial anastomoses performed adjacent to a bifurcation may have a high incidence of thrombosis due to turbulence at the bifurcation (O’Brien, 1977). This problem was investigated by anastomising rabbit femoral arteries close (less than 2 mm) to the bifurcation of the artery into the deep femoral artery and saphena artery. The patency rates could then be compared with extensive previous series of femoral arterial repairs carried out in this Unit. An alternative arterial reconstruction utilising a Y-shaped microvascular graft with anastomoses distant from the bifurcation was also studied. MATERIALSANDMETHOD Thirty adult New Zealand white rabbits of both sexes weighing an average of 2 kg were used. Anaesthesia was induced with 30 mg of intravenous pentobarbitone sodium, and maintained with nitrous oxide, halothane and oxygen via a mask. Incisions were made in both groins, and the femoral artery and its bifurcation into the deep femoral and saphena arteries were exposed (Barone, 1973). On one side the femoral artery was divided 1-2 mm proximal to the bifurcation and repaired (Fig. Ia). On the second side the femoral artery was divided I cm proximal to the bifurcation, the saphena artery I cm distal from the bifurcation, and the deep femoral just short of I cm Deep FemoralA
.::. ,I :T’j-: 1 cm. ,‘.,
.,.:_:.. :.::..:; Gut’
C- Femoral A
$--Femoral
+ FIG.
IA.
FIG.
Anatomy IB.
A
of anastomosis proximal to bifurcation.
Anatomy of Y-shaped
Address for reprints : B. McC. O’Brien, Director, Melbourne, 3065, Australia. 158
microarterial
Microsurgery
graft.
Research Unit, St Vincent’s Hospital,
Y-SHAPED MICROARTERIALGRAFTS
I59
The Y-shaped arterial segment was used as an orthotopic distal from the bifurcation. microarterial graft to re-establish arterial blood flow (Fig. Ib). Uniform microvascular techniques were used for all anastomoses. After exposing the vessels, spasm was relieved by topical 2 per cent xylocaine. The vessels were placed in a double approximator clamp and irrigated with heparinised saline. Interrupted IO/O nylon sutures (Ethicon BV-2 needle) were used. Patency was determined immediately after completing the anastomosis. The anastomoses were marked with 4/o black silk sutures, and the skin closed with interrupted 4/o nylon sutures. No systemic anticoagulants were used. One week later the rabbits were explored under general anaesthesia and the patency tested with No. 2 jewellers’ forceps. Eleven animals were perfused at that time for scanning electron microscopy studies; the other animals were sacrificed. Vessels to be studied with scanning electron microscopy were dilated with 2 per cent topical Xylocaine. The abdominal aorta was canulated and perfused with normal saline for 5 minutes. The femoral arterial system was then perfused with Karnovsky’s fixative diluted I : 3 with O’IM sodium cacodylate buffer, pH 7.3. The vena cava was divided to allow for egress of the solution. Perfusion was performed for 15 minutes in situ. The vessels were then excised, and fixation continued in Karnovsky’s fixative diluted I : I in the same buffer, at 4°C. The vessels were rinsed 3 times in cacodylate buffer, then post fixed in 2 per cent osmium tetroxide for 2 hours. They were then rinsed in distilled water, dehydrated and immersed in propylene oxide for 5 minutes. The tissue was transferred to a I : I solution of propylene oxide and camphene for 30 minutes at 45C. The vessels were then placed in pure camphene for an additional 30 minutes and They were then divided longitudinally, mounted, then dried by vacuum sublimation. and coated with gold. The specimens were examined in a Joel SM-35 scanning electron microscope at an accelerating voltage of 25 KV. RESULTS The patency rate for the 30 anastomoses 1-2 mm proximal to the bifurcation IOO per cent at operation and at I week.
FIG. z. Results of Y-shaped microarterial graft study.
was
BRITISH
160
JOURNAL
OF
PLASTIC
SURGERY
The patency rate for the go anastomoses of the 30 Y-shaped microarterial grafts was 97 per cent at surgery. One saphena and two deep femoral arteries were thrombosed. At one week the patency rate for the go anastomoses was 86 per cent. Two femoral, 8 deep femoral and 3 saphena arteries were thrombosed. In 28 of the 30 Y-shaped grafts, at least I branch was patent, a “flow patency” rate of g3 per cent (Fig. 2).
FIG. 3~.
Scanning electron micrograph of rabbit femoral, saphena and deep femoral system showing endothelialised surface. At one week. ( x 36). (Line bottom right = IOOO microns)
FIG. 3B. microvilli
Anastomotic site of same vessel. and clearly defined cell junctions.
FIG. 3c.
Bifurcation
FIG. 4.
Fibrin
IO/O nylon sutures covered with endothelium with normal At one week. ( x 80). (Line bottom right = IOO microns)
of same vessel with flow oriented endothelium. Note clean surface predicted turbulence. At one week. ( x 65).
Erythrocytes film extending from anastomotic site towards bifurcation. in the mesh. ( x 55). (Line bottom right = IOO microns)
at points of
are seen trapped
Y-SHAPED MICROARTERIALGRAFTS
161
Eleven patent anastomoses proximal to bifurcations were studied with scanning All vessels had normal endothelium with no thrombus proximal electron microscopy. to the anastomotic site. Eight specimens (73 per cent) had almost complete endothelialisation at the anastomotic site, at the bifurcation and in the distal branches with variable coverage with scattered erythrocytes and platelets at these sites. There was no relationship of the position of the deposits to predicted areas of flow turbulence (Figs. 3a, b, c). One specimen had a fine fibrin net extending from the anastomotic site to the point of bifurcation. A few erythrocytes were trapped in the mesh, but this area was otherwise free of thrombus. Distally the branches had a moderate film of platelets, erythrocytes and fibrin. This area was not related to turbulence and did not affect flow (Fig. 4). Two specimens had a thin layer of thrombotic build up at the anastomotic site and immediately distal to it, consisting of erythrocytes and some platelets. The position of the thrombosis was adjacent to the ostia of the branches, a predicted area of turbulence. However, it should be noted that this was a relatively thin film deposit that would be unlikely to affect patency. In the distal branches of these two specimens approximately 85 per cent of the surface area was covered with erythrocytes and platelets. In one of these specimens the deposits appeared capable of affecting flow, although the patency test was positive. DISCUSSION Areas of turbulence occur when fluid flows through a bifurcation. Platelet aggregates have been demonstrated opposite the ostia of intercostal arteries on normal aortic Blood cells may be traumatised in the turbulence of the vortex and endothelium. liberate ADP, which will further stimulate platelet adherence to vessel walls (Jorgensen, Haerem and Moe, 1973; Spaet and Gaynor, 1970). Microvascular repairs are rapidly covered with a film of platelets. The cumulative platelet aggregation of these effects are believed by many investigators to increase the risk of thrombosis of microvascular anastomoses performed adjacent to bifurcations. As a result of these theoretical arguments and of a number of circumstantial clinical events, controversy regarding arterial reconstruction immediately proximal to a bifurcation has been generated. Three reconstructive techniques have been advocated : a single anastomosis proximal to the bifurcation; resection of the bifurcation and replacement with a Y-shaped graft with anastomosis some distance from the bifurcation; resection of the bifurcation, ligation of the minor branch and a single end to end anastomosis to the major branch. No previous attempt has been made to clarify this aspect of reconstructive microvascular surgery. The rabbit femoral artery bifurcation into the deep femoral and saphena arteries was used in this study as this unit had adopted this system as its basic microvascular model. Hayhurst and O’Brien (1975) reported a long term patency rate of 98 per cent for 50 end to end anastomoses of rabbit femoral arteries performed in the same fashion as in this study. Analysis reveals that at the 95 per cent confidence level there is no statistical difference between the patency rates of end to end anastomoses and anastomoses proximal to a bifurcation. The flow patency rate for the Y-shaped microarterial grafts is 93 per cent.
162
BRITISH
ARTERIAL
JOURNAL OF PLASTIC
SURGERY
Arterial patency at one week
RECONSTRUCTION
bifurcation, Y-shaped microarterial graft
S No statistical
difference
at 95% level of confidence.
FIG. 5. Patency rates for arterial reconstruction
at bifurcations.
Statistical analysis again fails to reveal any difference at the 95 per cent confidence level between the patency rates of the single anastomoses proximal to the bifurcation and that of the Y-shaped microarterial grafts (Fig. 5). The greater number of non-patent deep femoral anastomoses (8) than saphena anastomoses (3) was considered to be due to the dilhculty in dissection and the awkward The saphena artery, in spite of its smaller size (0.5 mm), position of the anastomoses. had a higher patency rate. Scanning electron microscopy revealed the majority (73 per cent) of the vessels to be relatively free of thrombi, and healing at rates comparable to a simple end to end repair. The average percentage of coverage of luminal surface for this series was 20 per cent. Previous work in this unit showed an average coverage of luminal surface in simple end to end repairs of 40 per cent at one week (Nightingale et al., 1978). A major difference in the 2 series was a preponderance of fibrin and erythrocyte thrombus in the bifurcation series, compared to platelet thrombus in the simple end to end series. In the simple end to end series the platelets were evenly divided between flattened and active forms. In the bifurcation series the platelets were nearly all of the active form (Walsh and Barnhard, 1973). The position of the thrombus in the bifurcation series was related to flow turbulence in only 2 cases. In the other specimens it was related to clamp damage, handling and the suture line. In light of the experimental findings, it would seem that the technically simplest arterial reconstruction for anastomoses proximal to bifurcations should be used and in most cases this would be an end to end anastomosis. SUMMARY The bifurcation of the rabbit femoral artery was used as a model to study the patency rates of arterial anastomoses performed immediately proximal to bifurcations Alternative arterial reconstructions utilising Y-shaped microarterial grafts were studied, Deleterious and data from previous work on end to end repairs was used for comparison.
Y-SHAPED
MICROARTERIAL
GRAFTS
163
effects of turbulence at bifurcations have not been demonstrated as there is no statistical difference in the patency rates of anastomoses proximal to bifurcations, end to end anastomoses distant from branching, and Y-shaped microarterial grafts. The assistance of Miss N. Carroll with the scanning electron microscopy is gratefully acknowledged. This project was supported by the National Health and Medical Research Council of Australia. REFERENCES
BARONE, R. (1973). “Atlas d’Anatomie du Lapin”. Paris: Masson, p. 131. HAYHURST? J. W. and O’BRIEN,B. McC. (1975). An experimental study of microvascular techmque, patency rates and related factors. BritishJournal of Plastic Surgery, 28, 128. JORGENSEN, L., HAEREM, J. W. and MOE, N. (1973). Platelet thrombosis and non-traumatic intimal injury in mouse aorta. Thrombosiset Diathesis Haemorrhagica, 29,470. NIGHTINGALE, GEMMA,FOGDESTAM, I. and O’BRIEN,B. McC. (1978). In preparation. O’BRIEN,B. McC. (1977). “Microvascular Reconstructive Surgery”. Edinburgh: Churchill Livingstone, p. 46. WALSH,R. T. and BARNHART, M. I. (x973). Blood platelet surfaces in 3-Dimension. SEM-IIRTI, pp. 481-488.
ofPlastic
Surgery (1979),
JZ, 158-163
PATENCY OF ANASTOMOSES ADJACENT TO THE BIFURCATION OF THE RABBIT FEMORAL ARTERY AND A COMPARISON WITH Y-SHAPED MICROARTERIAL GRAlW3 By D. De La PAVA, M.D., GEMMANIGHTINGALE, M.Sc.(Melb), B. B. SHAFIROFF, M.D. and B. McC. O’BRIEN,B.Sc., M.D., MS., F.R.C.S., F.R.A.C.S., F.A.C.S. Microsurgery Research Unit, St Vincent’s Hospital, Melbourne, 3065, Australia
THE major cause of failure in replantation and microvascular free tissue transfer surgery in our unit has been arterial thrombosis and a number of causative factors have been identified. It has been suggested that arterial anastomoses performed adjacent to a bifurcation may have a high incidence of thrombosis due to turbulence at the bifurcation (O’Brien, 1977). This problem was investigated by anastomising rabbit femoral arteries close (less than 2 mm) to the bifurcation of the artery into the deep femoral artery and saphena artery. The patency rates could then be compared with extensive previous series of femoral arterial repairs carried out in this Unit. An alternative arterial reconstruction utilising a Y-shaped microvascular graft with anastomoses distant from the bifurcation was also studied. MATERIALSANDMETHOD Thirty adult New Zealand white rabbits of both sexes weighing an average of 2 kg were used. Anaesthesia was induced with 30 mg of intravenous pentobarbitone sodium, and maintained with nitrous oxide, halothane and oxygen via a mask. Incisions were made in both groins, and the femoral artery and its bifurcation into the deep femoral and saphena arteries were exposed (Barone, 1973). On one side the femoral artery was divided 1-2 mm proximal to the bifurcation and repaired (Fig. Ia). On the second side the femoral artery was divided I cm proximal to the bifurcation, the saphena artery I cm distal from the bifurcation, and the deep femoral just short of I cm Deep FemoralA
.::. ,I :T’j-: 1 cm. ,‘.,
.,.:_:.. :.::..:; Gut’
C- Femoral A
$--Femoral
+ FIG.
IA.
FIG.
Anatomy IB.
A
of anastomosis proximal to bifurcation.
Anatomy of Y-shaped
Address for reprints : B. McC. O’Brien, Director, Melbourne, 3065, Australia. 158
microarterial
Microsurgery
graft.
Research Unit, St Vincent’s Hospital,
Y-SHAPED MICROARTERIALGRAFTS
I59
The Y-shaped arterial segment was used as an orthotopic distal from the bifurcation. microarterial graft to re-establish arterial blood flow (Fig. Ib). Uniform microvascular techniques were used for all anastomoses. After exposing the vessels, spasm was relieved by topical 2 per cent xylocaine. The vessels were placed in a double approximator clamp and irrigated with heparinised saline. Interrupted IO/O nylon sutures (Ethicon BV-2 needle) were used. Patency was determined immediately after completing the anastomosis. The anastomoses were marked with 4/o black silk sutures, and the skin closed with interrupted 4/o nylon sutures. No systemic anticoagulants were used. One week later the rabbits were explored under general anaesthesia and the patency tested with No. 2 jewellers’ forceps. Eleven animals were perfused at that time for scanning electron microscopy studies; the other animals were sacrificed. Vessels to be studied with scanning electron microscopy were dilated with 2 per cent topical Xylocaine. The abdominal aorta was canulated and perfused with normal saline for 5 minutes. The femoral arterial system was then perfused with Karnovsky’s fixative diluted I : 3 with O’IM sodium cacodylate buffer, pH 7.3. The vena cava was divided to allow for egress of the solution. Perfusion was performed for 15 minutes in situ. The vessels were then excised, and fixation continued in Karnovsky’s fixative diluted I : I in the same buffer, at 4°C. The vessels were rinsed 3 times in cacodylate buffer, then post fixed in 2 per cent osmium tetroxide for 2 hours. They were then rinsed in distilled water, dehydrated and immersed in propylene oxide for 5 minutes. The tissue was transferred to a I : I solution of propylene oxide and camphene for 30 minutes at 45C. The vessels were then placed in pure camphene for an additional 30 minutes and They were then divided longitudinally, mounted, then dried by vacuum sublimation. and coated with gold. The specimens were examined in a Joel SM-35 scanning electron microscope at an accelerating voltage of 25 KV. RESULTS The patency rate for the 30 anastomoses 1-2 mm proximal to the bifurcation IOO per cent at operation and at I week.
FIG. z. Results of Y-shaped microarterial graft study.
was
BRITISH
160
JOURNAL
OF
PLASTIC
SURGERY
The patency rate for the go anastomoses of the 30 Y-shaped microarterial grafts was 97 per cent at surgery. One saphena and two deep femoral arteries were thrombosed. At one week the patency rate for the go anastomoses was 86 per cent. Two femoral, 8 deep femoral and 3 saphena arteries were thrombosed. In 28 of the 30 Y-shaped grafts, at least I branch was patent, a “flow patency” rate of g3 per cent (Fig. 2).
FIG. 3~.
Scanning electron micrograph of rabbit femoral, saphena and deep femoral system showing endothelialised surface. At one week. ( x 36). (Line bottom right = IOOO microns)
FIG. 3B. microvilli
Anastomotic site of same vessel. and clearly defined cell junctions.
FIG. 3c.
Bifurcation
FIG. 4.
Fibrin
IO/O nylon sutures covered with endothelium with normal At one week. ( x 80). (Line bottom right = IOO microns)
of same vessel with flow oriented endothelium. Note clean surface predicted turbulence. At one week. ( x 65).
Erythrocytes film extending from anastomotic site towards bifurcation. in the mesh. ( x 55). (Line bottom right = IOO microns)
at points of
are seen trapped
Y-SHAPED MICROARTERIALGRAFTS
161
Eleven patent anastomoses proximal to bifurcations were studied with scanning All vessels had normal endothelium with no thrombus proximal electron microscopy. to the anastomotic site. Eight specimens (73 per cent) had almost complete endothelialisation at the anastomotic site, at the bifurcation and in the distal branches with variable coverage with scattered erythrocytes and platelets at these sites. There was no relationship of the position of the deposits to predicted areas of flow turbulence (Figs. 3a, b, c). One specimen had a fine fibrin net extending from the anastomotic site to the point of bifurcation. A few erythrocytes were trapped in the mesh, but this area was otherwise free of thrombus. Distally the branches had a moderate film of platelets, erythrocytes and fibrin. This area was not related to turbulence and did not affect flow (Fig. 4). Two specimens had a thin layer of thrombotic build up at the anastomotic site and immediately distal to it, consisting of erythrocytes and some platelets. The position of the thrombosis was adjacent to the ostia of the branches, a predicted area of turbulence. However, it should be noted that this was a relatively thin film deposit that would be unlikely to affect patency. In the distal branches of these two specimens approximately 85 per cent of the surface area was covered with erythrocytes and platelets. In one of these specimens the deposits appeared capable of affecting flow, although the patency test was positive. DISCUSSION Areas of turbulence occur when fluid flows through a bifurcation. Platelet aggregates have been demonstrated opposite the ostia of intercostal arteries on normal aortic Blood cells may be traumatised in the turbulence of the vortex and endothelium. liberate ADP, which will further stimulate platelet adherence to vessel walls (Jorgensen, Haerem and Moe, 1973; Spaet and Gaynor, 1970). Microvascular repairs are rapidly covered with a film of platelets. The cumulative platelet aggregation of these effects are believed by many investigators to increase the risk of thrombosis of microvascular anastomoses performed adjacent to bifurcations. As a result of these theoretical arguments and of a number of circumstantial clinical events, controversy regarding arterial reconstruction immediately proximal to a bifurcation has been generated. Three reconstructive techniques have been advocated : a single anastomosis proximal to the bifurcation; resection of the bifurcation and replacement with a Y-shaped graft with anastomosis some distance from the bifurcation; resection of the bifurcation, ligation of the minor branch and a single end to end anastomosis to the major branch. No previous attempt has been made to clarify this aspect of reconstructive microvascular surgery. The rabbit femoral artery bifurcation into the deep femoral and saphena arteries was used in this study as this unit had adopted this system as its basic microvascular model. Hayhurst and O’Brien (1975) reported a long term patency rate of 98 per cent for 50 end to end anastomoses of rabbit femoral arteries performed in the same fashion as in this study. Analysis reveals that at the 95 per cent confidence level there is no statistical difference between the patency rates of end to end anastomoses and anastomoses proximal to a bifurcation. The flow patency rate for the Y-shaped microarterial grafts is 93 per cent.
162
BRITISH
ARTERIAL
JOURNAL OF PLASTIC
SURGERY
Arterial patency at one week
RECONSTRUCTION
bifurcation, Y-shaped microarterial graft
S No statistical
difference
at 95% level of confidence.
FIG. 5. Patency rates for arterial reconstruction
at bifurcations.
Statistical analysis again fails to reveal any difference at the 95 per cent confidence level between the patency rates of the single anastomoses proximal to the bifurcation and that of the Y-shaped microarterial grafts (Fig. 5). The greater number of non-patent deep femoral anastomoses (8) than saphena anastomoses (3) was considered to be due to the dilhculty in dissection and the awkward The saphena artery, in spite of its smaller size (0.5 mm), position of the anastomoses. had a higher patency rate. Scanning electron microscopy revealed the majority (73 per cent) of the vessels to be relatively free of thrombi, and healing at rates comparable to a simple end to end repair. The average percentage of coverage of luminal surface for this series was 20 per cent. Previous work in this unit showed an average coverage of luminal surface in simple end to end repairs of 40 per cent at one week (Nightingale et al., 1978). A major difference in the 2 series was a preponderance of fibrin and erythrocyte thrombus in the bifurcation series, compared to platelet thrombus in the simple end to end series. In the simple end to end series the platelets were evenly divided between flattened and active forms. In the bifurcation series the platelets were nearly all of the active form (Walsh and Barnhard, 1973). The position of the thrombus in the bifurcation series was related to flow turbulence in only 2 cases. In the other specimens it was related to clamp damage, handling and the suture line. In light of the experimental findings, it would seem that the technically simplest arterial reconstruction for anastomoses proximal to bifurcations should be used and in most cases this would be an end to end anastomosis. SUMMARY The bifurcation of the rabbit femoral artery was used as a model to study the patency rates of arterial anastomoses performed immediately proximal to bifurcations Alternative arterial reconstructions utilising Y-shaped microarterial grafts were studied, Deleterious and data from previous work on end to end repairs was used for comparison.
Y-SHAPED
MICROARTERIAL
GRAFTS
163
effects of turbulence at bifurcations have not been demonstrated as there is no statistical difference in the patency rates of anastomoses proximal to bifurcations, end to end anastomoses distant from branching, and Y-shaped microarterial grafts. The assistance of Miss N. Carroll with the scanning electron microscopy is gratefully acknowledged. This project was supported by the National Health and Medical Research Council of Australia. REFERENCES
BARONE, R. (1973). “Atlas d’Anatomie du Lapin”. Paris: Masson, p. 131. HAYHURST? J. W. and O’BRIEN,B. McC. (1975). An experimental study of microvascular techmque, patency rates and related factors. BritishJournal of Plastic Surgery, 28, 128. JORGENSEN, L., HAEREM, J. W. and MOE, N. (1973). Platelet thrombosis and non-traumatic intimal injury in mouse aorta. Thrombosiset Diathesis Haemorrhagica, 29,470. NIGHTINGALE, GEMMA,FOGDESTAM, I. and O’BRIEN,B. McC. (1978). In preparation. O’BRIEN,B. McC. (1977). “Microvascular Reconstructive Surgery”. Edinburgh: Churchill Livingstone, p. 46. WALSH,R. T. and BARNHART, M. I. (x973). Blood platelet surfaces in 3-Dimension. SEM-IIRTI, pp. 481-488.
