- Email: [email protected]
An implanted ultrasound doppler probe for microvascular monitoring: an experimental study
Brilish Journal of Plastic Surgery (1986) 39, I 18-l24 Q 1986 The Trustees of British Association of Plastic Surgeons
An implanted ultrasound Doppler probe for microvascular monitoring: an experimental study J. 0. ROBERTS, B. M. JONES and R. M. GREENHALGH Charing
Cross and Westminster
Medical
School,
London
Summary-An experimental investigation is presented into the use of pulsed ultrasound Doppler flowmetry with a percutaneously implanted probe as a monitor of microvascular anastomotic patency. The method accurately indicated and distinguished between experimental arterial and venous occlusion in epigastric island flaps. In free flaps it had advantages over manipulative, intraoperative tests of anastomotic patency and was a reliable post-operative monitor of flap circulation in the experimental model. The technique appears to have great potential as a post-operative monitor for free flaps, including those without a visible surface.
The survival of vascularised free tissue transfers depends on the patency of microvascular anastomoses. Ultrasound Doppler flowmetry reliably proves patency at surgery (Jones and Greenhalgh, 1983) but post-operative occulsion may occur as a result of external pressure, oedema or kinking of the pedicle. Of 2233 free flaps performed in the West and Japan up to 1981, 90 developed complications and were salvaged by re-operation (Shaw, 1981). The likelihood of flap survival is directly related to the speed with which circulatory embarrassment is recognised and corrected (Goodstein and Buncke, 1979). The need for a reliable, continuous post-operative monitor is evident. Though several techniques are available for monitoring cutaneous free flaps (Jones, 1984), those without a visible surface cannot be monitored after wound closure. Transcutaneous ultrasound Doppler tlowmetry has been unsatisfactory as a post-operative monitor of microvascular anastomotic patency due to difficulty in correctly orientating the probe over the vessels (Karakowski and Buncke, 1975). In an experimental study a totally implanted, pulsed ultrasound Doppler flowmeter was effective in measuring portal vein blood flow in dogs over periods of up to 15 months (Anderson, 1981). This technique is inappropriate for clinical practice as surgery is required for removal of the probe. Work supported by a Locally Organised Research North West Thames Regional Health Authority.
Grant
from
118
A continuous wave ultrasound Doppler microprobe. diameter 2.3 mm, is valuable as an aid to locating vessels in coronary artery surgery (Moulder et al., 1977). A probe of such size could be implanted percutaneously, adjacent to the pedicle vessels, at surgery and withdrawn, like a drainage tube, at the end of the monitoring period. Continuous wave ultrasound Doppler has disadvantages when the probe is applied directly to vessel walls. As the transducer is composed of two crystals, one transmitting and one receiving, measurements can only be made within the zone of overlap of their fields. Immediately in front of the probe is a “dead area” of no overlap from which measurements cannot be recorded. The extent of the dead area is affected by the operating frequency of the equipment; if 8 MHz the area extends to 2.5 mm from the face of the probe and if 4 MHz to 15 mm. A further drawback is that backscattered ultrasound from anywhere in the transmitted beam, whatever the depth, will affect the output signal and produce significant errors when vessels are close together. Both of these disadvantages are overcome by using a pulsed ultrasound Doppler flowmeter. In this there is only one crystal, which alternately transmits and receives, so avoiding any dead area. The depth at which measurement is made is determined by the time interval between the two functions. This can be adjusted to target onto the vessel being monitored and the signal output will not be affected by flow in underlying or adjacent vessels.
AN IMPLANTED
ULTRASOUND
DOPPLER
PROBE
FOR MICROVASCULAR
119
MONITORING
Fig. 1 Figure
l~impiantable
ultrasound
Doppler
probe (size comparison
A pulsed ultrasound Doppler flowmeter with an implantable probe may have potential as a postoperative monitor of microvascular anastomotic patency in free flaps, including those without a visible surface. This study was undertaken to evaluate such an application of the equipment in experimental animals. Materials and methods The equipment consisted of a sterilisable probe, I cm long and 2 mm diameter (Fig. I), connected by a flexible, silastic sheathed wire to a 20MHz
EME
with a match stick).
pulsed ultrasound Doppler flowmeter (MF20, Eden Medizinische Elektronik GmbH) (Fig. 2). In addition to an acoustic signal, this provided continuous digital display of mean flow velocity (cm/ set) and bar graph display of both mean and instantaneous velocity. The instrument was connected to a chart recorder (Kipp and Zonen BD9) with chart speed set at 2 mm/min and sensitivity at 5 V, providing an 8 mm deflection per 1 cmjsec velocity change. All experiments were carried out in anaesthetised 300 to 350gm CFY rats. Induction and mainten-
MF 20’.’ ii;,‘,:: /
-S..
Fig. 2 Figure 2--Pulsed
ultrasound
Doppler
flowmeter.
.
120
BRITISH
ante anaesthesia was inhalational with oxygen, nitrous oxide and halothane. The animals were sacrificed at the end of each experiment. The animals were divided into three groups to assess the equipment’s response to simple arterial occlusion, arterial and venous occlusion in epigastric island flaps and continuous monitoring of epigastric free flaps. Experimental vascular occlusion was produced, in Groups 1 and 2, by a loop of 3/O nylon passed percutaneously through a 20 gauge intravenous cannula. The ultrasound probe was inserted percutaneously and positioned, under direct vision, at approximately 45” to the artery which the transducer face touched but did not distort. Probe position was maintained by anchoring it to adjacent tissues with 8/O silk sutures. Coupling gel was not required, a drop of blood or saline sufficing. Depth gate setting was adjusted to give the maximum signal output for the vessel being monitored. All wounds were closed prior to monitoring and the probe withdrawn percutaneously at the end of each experiment. Group I
Simple arterial occlusion
The femoral vessels were exposed via a 3 cm oblique groin incision. An occlusion loop was placed around the superficial femoral artery and the ultrasound probe positioned 1 to 2cm distally (Fig. 3). Mean blood flow velocity in the superficial femoral artery was measured before, during and after release of occlusion.
JOURNAL
OF PLASTIC
Group 2 Arterial and venous occlusion in epigastric islandflaps
Island flaps measuring 5 x 4cm were raised based on the superficial epigastric vessels. All other branches of the femoral artery and vein were ligated so that all blood flowing through these vessels was directed through the flap. An arterial occlusion loop was placed around the superficial femoral artery and the ultrasound probe positioned 1 to 2cm distally. A second occlusion loop was placed around the femoral vein (Fig. 4). Mean and instantaneous arterial blood flow velocity was measured with the vessels unoccluded and in the presence of arterial and venous occlusion separately. Group 3 flaps
Colztinuous monitoring of epigastric free
Flaps were raised, ultrasound probe positioned and arterial blood velocity measured as in Group 2. The femoral vessels were then divided and anastomosed, in situ, by standard microvascular techniques. The probe was repositioned, if necessary, so that it was about 1 cm distal to the arterial anastomosis (Fig. 5). Flow velocity recording was begun at the time of clamp removal. Ten to 15 minutes later anastomotic patency was assessed by double forceps patency test (Hayhurst and O’Brien, 1975) raise test (Acland, 1972) and insonation and venous flow augmentation test (Jones and Greenhalgh, 1983). The flap was then sutured in place and monitored, under anaesthetic, for approximately 3 hours.
Pulsed uttrasound Pulsed ultrasound Occlusion
loop -
I Superficial femoral artery
Fig. 3 Figure 3-Simple
arterial
occlusion
Fig. 4 (Group
I).
SURGERY
Figure 4-Epigastric
island flap (Group
2)
AN IMPLANTED
ULTRASOUND
DOPPLER
PROBE
FOR MICROVASCULAR
121
MONITORING
,It
0
10
:
20
t Occlusion 1 Release of occlusion
40
50 Time (min
)
Fig. 6
Flgure &Trace of superficlal femoral artery mean blood velocity before. during and after release of occlusion (Group I).
Femoral vein blood flow was insonated and its velocity measured by altering the depth gate setting. Mean flow velocity was 0.95cm/sec (range 0.75-1.25 cmjsec).
Fig. 5 Figure 5-Epigastric
free flap (Group
3)
At the end of the experiment the probe was withdrawn and the flap re-raised. The vessels were examined for any evidence of trauma caused by probe removal and the anastomoses reassessed by the above tests and, ultimately, by division of the vessels distal to the arterial anastomosis and proximal to the venous. Results Group I (5 rats) The mean pre-occlusion arterial blood velocity was 8.5 cmjsec (range 8-9 cmjsec). Arterial occlusion produced an immediate fall in flow velocity to 0 on each occasion. Release of occlusion resulted in a rise in flow velocity to a mean steady level of 8 cm/ set (range 7.5~8.5cm/sec) in a mean time of 19 minutes (range 10-22 minutes) (Fig. 6). The unoccluded artery gave a typical ultrasound Doppler acoustic signal which ceased abruptly on . . occlusion.
Group 2 (I 0 rats) Mean, pre-occlusion arterial blood velocity was 4.78 cmjsec (range 3.25-6.5 cm/set). Arterial occlusion (5 minutes duration) produced an immediate fall in flow velocity to 0 and cessation of the acoustic signal on each occasion. Release of occlusion resulted in a rise in flow velocity to a mean steady level of 4.25cm/sec (range 1.5-6.5cm/sec) in a mean time of 26.3 minutes (range 1240 minutes). Venous occlusion (I 5 minutes duration) resulted in an immediate fall in arterial blood velocity to 0 and an alteration of the acoustic signal to a stacatto sound. Release of occlusion “damped”, produced a rise in flow velocity to a mean of 2.75 cmjsec (range 2-3.75 cm/set) in a mean time of 26.3 minutes (range 2-65 minutes) (Fig. 7). Instantaneous flow oscillated with the pulse wave. The mean magnitude of oscillation in instantaneous flow was equivalent to 1.75 cm/set in the pre-occlusion artery, 0 during arterial occlusion and 0.45 cmjsec during venous obstruction. Insonation of the femoral vein, by altering the depth gate setting, produced a characteristic acoustic signal but the flow velocity was too low to record. Group 3 (13 rats) Ten flaps were successful, _ . culation.
3 failed to develop
a cir-
122
BRITISH -
Instantaneous
=
Mean velocity
I
0
20
OF PLASTIC
SURGERY
velocity
,
10
JOURNAL
30
I
60
7
70
80
110 Time (min )
Fig. I Figure ‘T-Trace of arterial mean and instantaneous blood velocity in an epigastric island flap before, during and after release of arterial and venous occlusion (Group 2).
Fig. 8 Figure 8-Superficial
Assessment 10 minutes after clamp removal showed arterial and venous anastomoses to be patent, by all tests, in 12 of the 13 flaps. In the remaining flap the venous anastomqsis was patent but the arterial anastomosis occluded. Revision by excision of the occluded segment and direct reanastomosis again obstructed but a second revision, using an interpositional vein graft, remained patent and the flap was successful. In 3 flaps, despite patent anastomoses, the mean flow velocity remained 0 over a 3 hour monitoring period. Instantaneous flow velocity oscillation persisted throughout (mean magnitude of oscillation 0.48 cmjsec). When the flaps were re-raised after 3 hours both anastomoses were patent but the superficial epigastric artery wasin spasm preventing any forward flow through the flap (Fig. 8). Among the successful flaps there was considerable variation in flow rate (Fig. 9). However, in all but one, the flow velocity slowly increased to reach a final steady level with any downward fluctuations being short lived and never reaching 0. In the one exception there was early very high flow which fell to a steady level of 5.5 cmjsec by 2 hours. The final flow velocity in the successful flapsmean 4.5cm/sec, range 3.25 to 5.5cm/sec-was greater than that in the vessels before division and anastomosis-mean 3.25 cm/set, range 2-5.75 cm/ sec. The acoustic signal in all the free flaps was initially the “damped” sound heard on venous occlusion in Group 2, instantaneous velocity oscillation was comparably low with a mean of 0.45 cm/ sec. In the 3 flaps with superficial epigastric artery
epigastric
artery spasm.
spasm the “damped” sound and low instantaneous oscillation persisted throughout. In the 10 successful flaps the sound changed to the smooth “whoosh” characteristic of a freely flowing artery and the instantaneous velocity oscillation increased in magnitude to a mean of 1.28 cmjsec. When the flaps were elevated at the end of each experiment, inspection of the vessels and surrounding tissues revealed no evidence of trauma resulting from percutaneous probe withdrawal. All anastomoses were patent on testing and division. Discussion Ultrasound
Doppler
flowmetry
is of proven
value
‘Olf-----l 0
30
60
120
150
180
Time (min 1
Fig. 9 Figure 9-Arterial mean blood velocity in epigastric free flaps. Recordings in 9 of the 10 flaps fell within the shaded area; the exception, where there was unusually high early flow, is shown by the separate upper line (Group 3).
AN IMPLANTED
ULTRASOUND
DOPPLER
PROBE
FOR MICROVASCULAR
in per-operative anastomotic assessment in both large vessel surgery (Keitzer et al., 1972) and microvascular free tissue transfers (Jones, 1985). The advantages of pulsed, rather than continuous, wave form equipment, when applied per-operatively to vessels 1 mm or less in diameter, has been demonstrated in neurosurgery (Gilsbach, 1983). The drawbacks of ultrasound Doppler flowmetry as a transcutaneous post-operative monitor (Karkowski and Buncke, 1975) could, perhaps, be overcome by using an implanted transducer positioned directly against the pedicle artery. The small, sterilisable probe and pulsed wave form of the equipment used in this study make it potentially suitable for such an approach. It has a digital display, which simplifies interpretation for relatively inexperienced observers, and a chart recorder provides a permanent record. Blood velocity is expressed in cmjsec. rather than the more correct unit of kilohertz of Doppler shift. The Doppler shift frequency is expressed by the formulaAF = 2.f.v.cos $J C
where AF is the frequency shift, f the frequency of the incident ultrasound, 4 the angle between the incident beam and the line of movement (i.e. blood flow) and c the velocity of sound in the tissue. Thus the frequency shift is affected by the incident angle (i.e. that between the probe and the blood vessel). This equipment is calibrated assuming the angle to be 45”. It must be appreciated that this is impossible to control accurately and so the flow velocity should not be considered an absolute value. The experimental model was chosen as the epigastric flap in the rat has been used extensively in studying monitoring techniques. As the proposed monitor directly examines blood flow through microanastomoses the presence of a visible skin surface is irrelevant to the results. The equipment reliably and rapidly indicated experimental arterial occlusion in the superficial femoral artery and arterial and venous obstruction in epigastric island flaps by an immediate fall in arterial blood velocity to 0. In addition, arterial and venous occlusion could be distinguished. In the former the Doppler sound and instantaneous oscillation ceased. In contrast, venous occlusion produced a “damped” staccato sound and a reduced, but persisting instantaneous flow oscillation. In the free flaps, superficial epigastric artery spasm, with absent flap circulation, was detected by ultrasound Doppler flowmetry but not by the double forceps or raise tests. Although this spasm
123
MONITORING
may resolve, flap survival could not be relied upon. This was dramatically illustrated in one instance where spasm was found when the flap was raised after 1 hour but when the vessels were examined at the end of the experiment (2 hours later) the proximal superficial epigastric artery, which was not in spasm, had become occluded by clot although the anastomoses remained patent. A11 the successful flaps demonstrated the same pattern of mean arterial blood velocity and magnitude of instantaneous oscillation increasing until a steady level was reached. Late occlusion did not occur but early arterial obstruction was clearly indicated. In two flaps, in an attempt to mimic late failure, a solution of adrenaline (1:200 000) was irrigated beneath the flap after 3 hours monitoring. Constriction of small vessels produced a rapid fall in mean flow velocity to 0 on both occasions. Longer flap monitoring, after the animals had been recovered from anaesthetic, would have been desirable but was impractical in this model. Clinical evaluation is in progress and will be the subject of a later report. Conclusion
Pulsed ultrasound Doppler flowmetry with an implanted probe was valuable as a per-operative indicator of anastomotic status and showed great potential as a continuous postoperative monitor of vascularised free tissue transfers in an experimental model. Acknowledgement The ultrasound Doppler caid Limited. Chichester.
flowmeter
was kindly loaned
by Soni-
References Acland, R. D. (1972). Signs of patency in small vessel anastomosis. Surgery. 72.744. Anderson, M. F. (1981). Pulsed Doppler ultrasomc flowmeter: Application to the study of hepatic blood flow. In: Granger, D. N.. and Bulklev. G. B. (Eds) Measurement nfEIood FIoM’. Applications io the Splanchnic kirculation. p. 345.Baltimore. Williams and Wilkins. Gilsbach, J. M. (1983). lntraoperative Doppler Sonography in Neurosurger?. Vienna, Springer-Verlag. Goodstein, W. A. and Buncke, H. J. Jr. (1979). Patterns of vascular anastomosis vs success of free groin flap transfers. Plastic and Reconstructive Surgery, 64, 37. Hayhurst, J. W. and O’Brien, B.Mc. (1975). An experimental study of microvascular technique, patency rates and related factors. British Journal ofplastic Surgery, 28, 128. Jones, B. M. and Greenhalgh, R. M. (1983). The use of the ultrasound Doppler flowmeter in reconstructive microvascular surgery. Britkh Journal of Plastic Surgery, 36,245.
124 Jones, B. M. (1984). Monitors for the cutaneous microcirculation. Plastic and Reconstructive Surgery, 73,843. Jones, B. M. (1985). Predicting the fate of free tissue transfers. Annals of the Royal College of Surgeons of England, 67,63. Karkowski, J. and Buncke, H. J. Jr. (1975). A simplified technique for free transfer of groin flaps by use of a Doppler probe. Plastic and Reconstructive Surgery, 55,682. Keitzer, W. F., Licht, E. L., Brossart, F. A. and De Weese, M. S. (1972). Use of the Doppler ultrasonic flowmeter during arterial vascular surgery. &chives ofSurgery, 105,308. _ Moulder. P. V.. Teaew. _, M. J.., Manuele. V. J.., Brunswick. R. A. and Daicoff, G. R. (1977). Intraoperative Doppler coronary artery finder. Annals of Thoracic Surgery, 24.430. Shaw, W. (1981). A general survey of free flaps. Unpublished work presented at “Clinical Frontiers in Reconstructive Microsurgery”, 26th June, 1981,
BRITISH
JOURNAL
OF PLASTIC
SURGERY
The Authors J. 0. Roberts, MA, FRCS, Lecturer in Surgery, - _ Charing Cross and Westminster Medical School, London, B. M. Jones. MS. FRCS. Consultant Plastic Surgeon. The Hospital for Sick dhildren, Great Ormond Street&d University College Hospital, London. R. M. Greenhalgh, MA, MD, MChir, FRCS, Professor of Surgery, Charing Cross and Westminster Medical School, London Requests for reprints to: J. 0. Roberts, ment of Surgery, Charing Cross Hospital, London W6 8RF.
MA, FRCS, DepartFulham Palace Road.
An implanted ultrasound Doppler probe for microvascular monitoring: an experimental study J. 0. ROBERTS, B. M. JONES and R. M. GREENHALGH Charing
Cross and Westminster
Medical
School,
London
Summary-An experimental investigation is presented into the use of pulsed ultrasound Doppler flowmetry with a percutaneously implanted probe as a monitor of microvascular anastomotic patency. The method accurately indicated and distinguished between experimental arterial and venous occlusion in epigastric island flaps. In free flaps it had advantages over manipulative, intraoperative tests of anastomotic patency and was a reliable post-operative monitor of flap circulation in the experimental model. The technique appears to have great potential as a post-operative monitor for free flaps, including those without a visible surface.
The survival of vascularised free tissue transfers depends on the patency of microvascular anastomoses. Ultrasound Doppler flowmetry reliably proves patency at surgery (Jones and Greenhalgh, 1983) but post-operative occulsion may occur as a result of external pressure, oedema or kinking of the pedicle. Of 2233 free flaps performed in the West and Japan up to 1981, 90 developed complications and were salvaged by re-operation (Shaw, 1981). The likelihood of flap survival is directly related to the speed with which circulatory embarrassment is recognised and corrected (Goodstein and Buncke, 1979). The need for a reliable, continuous post-operative monitor is evident. Though several techniques are available for monitoring cutaneous free flaps (Jones, 1984), those without a visible surface cannot be monitored after wound closure. Transcutaneous ultrasound Doppler tlowmetry has been unsatisfactory as a post-operative monitor of microvascular anastomotic patency due to difficulty in correctly orientating the probe over the vessels (Karakowski and Buncke, 1975). In an experimental study a totally implanted, pulsed ultrasound Doppler flowmeter was effective in measuring portal vein blood flow in dogs over periods of up to 15 months (Anderson, 1981). This technique is inappropriate for clinical practice as surgery is required for removal of the probe. Work supported by a Locally Organised Research North West Thames Regional Health Authority.
Grant
from
118
A continuous wave ultrasound Doppler microprobe. diameter 2.3 mm, is valuable as an aid to locating vessels in coronary artery surgery (Moulder et al., 1977). A probe of such size could be implanted percutaneously, adjacent to the pedicle vessels, at surgery and withdrawn, like a drainage tube, at the end of the monitoring period. Continuous wave ultrasound Doppler has disadvantages when the probe is applied directly to vessel walls. As the transducer is composed of two crystals, one transmitting and one receiving, measurements can only be made within the zone of overlap of their fields. Immediately in front of the probe is a “dead area” of no overlap from which measurements cannot be recorded. The extent of the dead area is affected by the operating frequency of the equipment; if 8 MHz the area extends to 2.5 mm from the face of the probe and if 4 MHz to 15 mm. A further drawback is that backscattered ultrasound from anywhere in the transmitted beam, whatever the depth, will affect the output signal and produce significant errors when vessels are close together. Both of these disadvantages are overcome by using a pulsed ultrasound Doppler flowmeter. In this there is only one crystal, which alternately transmits and receives, so avoiding any dead area. The depth at which measurement is made is determined by the time interval between the two functions. This can be adjusted to target onto the vessel being monitored and the signal output will not be affected by flow in underlying or adjacent vessels.
AN IMPLANTED
ULTRASOUND
DOPPLER
PROBE
FOR MICROVASCULAR
119
MONITORING
Fig. 1 Figure
l~impiantable
ultrasound
Doppler
probe (size comparison
A pulsed ultrasound Doppler flowmeter with an implantable probe may have potential as a postoperative monitor of microvascular anastomotic patency in free flaps, including those without a visible surface. This study was undertaken to evaluate such an application of the equipment in experimental animals. Materials and methods The equipment consisted of a sterilisable probe, I cm long and 2 mm diameter (Fig. I), connected by a flexible, silastic sheathed wire to a 20MHz
EME
with a match stick).
pulsed ultrasound Doppler flowmeter (MF20, Eden Medizinische Elektronik GmbH) (Fig. 2). In addition to an acoustic signal, this provided continuous digital display of mean flow velocity (cm/ set) and bar graph display of both mean and instantaneous velocity. The instrument was connected to a chart recorder (Kipp and Zonen BD9) with chart speed set at 2 mm/min and sensitivity at 5 V, providing an 8 mm deflection per 1 cmjsec velocity change. All experiments were carried out in anaesthetised 300 to 350gm CFY rats. Induction and mainten-
MF 20’.’ ii;,‘,:: /
-S..
Fig. 2 Figure 2--Pulsed
ultrasound
Doppler
flowmeter.
.
120
BRITISH
ante anaesthesia was inhalational with oxygen, nitrous oxide and halothane. The animals were sacrificed at the end of each experiment. The animals were divided into three groups to assess the equipment’s response to simple arterial occlusion, arterial and venous occlusion in epigastric island flaps and continuous monitoring of epigastric free flaps. Experimental vascular occlusion was produced, in Groups 1 and 2, by a loop of 3/O nylon passed percutaneously through a 20 gauge intravenous cannula. The ultrasound probe was inserted percutaneously and positioned, under direct vision, at approximately 45” to the artery which the transducer face touched but did not distort. Probe position was maintained by anchoring it to adjacent tissues with 8/O silk sutures. Coupling gel was not required, a drop of blood or saline sufficing. Depth gate setting was adjusted to give the maximum signal output for the vessel being monitored. All wounds were closed prior to monitoring and the probe withdrawn percutaneously at the end of each experiment. Group I
Simple arterial occlusion
The femoral vessels were exposed via a 3 cm oblique groin incision. An occlusion loop was placed around the superficial femoral artery and the ultrasound probe positioned 1 to 2cm distally (Fig. 3). Mean blood flow velocity in the superficial femoral artery was measured before, during and after release of occlusion.
JOURNAL
OF PLASTIC
Group 2 Arterial and venous occlusion in epigastric islandflaps
Island flaps measuring 5 x 4cm were raised based on the superficial epigastric vessels. All other branches of the femoral artery and vein were ligated so that all blood flowing through these vessels was directed through the flap. An arterial occlusion loop was placed around the superficial femoral artery and the ultrasound probe positioned 1 to 2cm distally. A second occlusion loop was placed around the femoral vein (Fig. 4). Mean and instantaneous arterial blood flow velocity was measured with the vessels unoccluded and in the presence of arterial and venous occlusion separately. Group 3 flaps
Colztinuous monitoring of epigastric free
Flaps were raised, ultrasound probe positioned and arterial blood velocity measured as in Group 2. The femoral vessels were then divided and anastomosed, in situ, by standard microvascular techniques. The probe was repositioned, if necessary, so that it was about 1 cm distal to the arterial anastomosis (Fig. 5). Flow velocity recording was begun at the time of clamp removal. Ten to 15 minutes later anastomotic patency was assessed by double forceps patency test (Hayhurst and O’Brien, 1975) raise test (Acland, 1972) and insonation and venous flow augmentation test (Jones and Greenhalgh, 1983). The flap was then sutured in place and monitored, under anaesthetic, for approximately 3 hours.
Pulsed uttrasound Pulsed ultrasound Occlusion
loop -
I Superficial femoral artery
Fig. 3 Figure 3-Simple
arterial
occlusion
Fig. 4 (Group
I).
SURGERY
Figure 4-Epigastric
island flap (Group
2)
AN IMPLANTED
ULTRASOUND
DOPPLER
PROBE
FOR MICROVASCULAR
121
MONITORING
,It
0
10
:
20
t Occlusion 1 Release of occlusion
40
50 Time (min
)
Fig. 6
Flgure &Trace of superficlal femoral artery mean blood velocity before. during and after release of occlusion (Group I).
Femoral vein blood flow was insonated and its velocity measured by altering the depth gate setting. Mean flow velocity was 0.95cm/sec (range 0.75-1.25 cmjsec).
Fig. 5 Figure 5-Epigastric
free flap (Group
3)
At the end of the experiment the probe was withdrawn and the flap re-raised. The vessels were examined for any evidence of trauma caused by probe removal and the anastomoses reassessed by the above tests and, ultimately, by division of the vessels distal to the arterial anastomosis and proximal to the venous. Results Group I (5 rats) The mean pre-occlusion arterial blood velocity was 8.5 cmjsec (range 8-9 cmjsec). Arterial occlusion produced an immediate fall in flow velocity to 0 on each occasion. Release of occlusion resulted in a rise in flow velocity to a mean steady level of 8 cm/ set (range 7.5~8.5cm/sec) in a mean time of 19 minutes (range 10-22 minutes) (Fig. 6). The unoccluded artery gave a typical ultrasound Doppler acoustic signal which ceased abruptly on . . occlusion.
Group 2 (I 0 rats) Mean, pre-occlusion arterial blood velocity was 4.78 cmjsec (range 3.25-6.5 cm/set). Arterial occlusion (5 minutes duration) produced an immediate fall in flow velocity to 0 and cessation of the acoustic signal on each occasion. Release of occlusion resulted in a rise in flow velocity to a mean steady level of 4.25cm/sec (range 1.5-6.5cm/sec) in a mean time of 26.3 minutes (range 1240 minutes). Venous occlusion (I 5 minutes duration) resulted in an immediate fall in arterial blood velocity to 0 and an alteration of the acoustic signal to a stacatto sound. Release of occlusion “damped”, produced a rise in flow velocity to a mean of 2.75 cmjsec (range 2-3.75 cm/set) in a mean time of 26.3 minutes (range 2-65 minutes) (Fig. 7). Instantaneous flow oscillated with the pulse wave. The mean magnitude of oscillation in instantaneous flow was equivalent to 1.75 cm/set in the pre-occlusion artery, 0 during arterial occlusion and 0.45 cmjsec during venous obstruction. Insonation of the femoral vein, by altering the depth gate setting, produced a characteristic acoustic signal but the flow velocity was too low to record. Group 3 (13 rats) Ten flaps were successful, _ . culation.
3 failed to develop
a cir-
122
BRITISH -
Instantaneous
=
Mean velocity
I
0
20
OF PLASTIC
SURGERY
velocity
,
10
JOURNAL
30
I
60
7
70
80
110 Time (min )
Fig. I Figure ‘T-Trace of arterial mean and instantaneous blood velocity in an epigastric island flap before, during and after release of arterial and venous occlusion (Group 2).
Fig. 8 Figure 8-Superficial
Assessment 10 minutes after clamp removal showed arterial and venous anastomoses to be patent, by all tests, in 12 of the 13 flaps. In the remaining flap the venous anastomqsis was patent but the arterial anastomosis occluded. Revision by excision of the occluded segment and direct reanastomosis again obstructed but a second revision, using an interpositional vein graft, remained patent and the flap was successful. In 3 flaps, despite patent anastomoses, the mean flow velocity remained 0 over a 3 hour monitoring period. Instantaneous flow velocity oscillation persisted throughout (mean magnitude of oscillation 0.48 cmjsec). When the flaps were re-raised after 3 hours both anastomoses were patent but the superficial epigastric artery wasin spasm preventing any forward flow through the flap (Fig. 8). Among the successful flaps there was considerable variation in flow rate (Fig. 9). However, in all but one, the flow velocity slowly increased to reach a final steady level with any downward fluctuations being short lived and never reaching 0. In the one exception there was early very high flow which fell to a steady level of 5.5 cmjsec by 2 hours. The final flow velocity in the successful flapsmean 4.5cm/sec, range 3.25 to 5.5cm/sec-was greater than that in the vessels before division and anastomosis-mean 3.25 cm/set, range 2-5.75 cm/ sec. The acoustic signal in all the free flaps was initially the “damped” sound heard on venous occlusion in Group 2, instantaneous velocity oscillation was comparably low with a mean of 0.45 cm/ sec. In the 3 flaps with superficial epigastric artery
epigastric
artery spasm.
spasm the “damped” sound and low instantaneous oscillation persisted throughout. In the 10 successful flaps the sound changed to the smooth “whoosh” characteristic of a freely flowing artery and the instantaneous velocity oscillation increased in magnitude to a mean of 1.28 cmjsec. When the flaps were elevated at the end of each experiment, inspection of the vessels and surrounding tissues revealed no evidence of trauma resulting from percutaneous probe withdrawal. All anastomoses were patent on testing and division. Discussion Ultrasound
Doppler
flowmetry
is of proven
value
‘Olf-----l 0
30
60
120
150
180
Time (min 1
Fig. 9 Figure 9-Arterial mean blood velocity in epigastric free flaps. Recordings in 9 of the 10 flaps fell within the shaded area; the exception, where there was unusually high early flow, is shown by the separate upper line (Group 3).
AN IMPLANTED
ULTRASOUND
DOPPLER
PROBE
FOR MICROVASCULAR
in per-operative anastomotic assessment in both large vessel surgery (Keitzer et al., 1972) and microvascular free tissue transfers (Jones, 1985). The advantages of pulsed, rather than continuous, wave form equipment, when applied per-operatively to vessels 1 mm or less in diameter, has been demonstrated in neurosurgery (Gilsbach, 1983). The drawbacks of ultrasound Doppler flowmetry as a transcutaneous post-operative monitor (Karkowski and Buncke, 1975) could, perhaps, be overcome by using an implanted transducer positioned directly against the pedicle artery. The small, sterilisable probe and pulsed wave form of the equipment used in this study make it potentially suitable for such an approach. It has a digital display, which simplifies interpretation for relatively inexperienced observers, and a chart recorder provides a permanent record. Blood velocity is expressed in cmjsec. rather than the more correct unit of kilohertz of Doppler shift. The Doppler shift frequency is expressed by the formulaAF = 2.f.v.cos $J C
where AF is the frequency shift, f the frequency of the incident ultrasound, 4 the angle between the incident beam and the line of movement (i.e. blood flow) and c the velocity of sound in the tissue. Thus the frequency shift is affected by the incident angle (i.e. that between the probe and the blood vessel). This equipment is calibrated assuming the angle to be 45”. It must be appreciated that this is impossible to control accurately and so the flow velocity should not be considered an absolute value. The experimental model was chosen as the epigastric flap in the rat has been used extensively in studying monitoring techniques. As the proposed monitor directly examines blood flow through microanastomoses the presence of a visible skin surface is irrelevant to the results. The equipment reliably and rapidly indicated experimental arterial occlusion in the superficial femoral artery and arterial and venous obstruction in epigastric island flaps by an immediate fall in arterial blood velocity to 0. In addition, arterial and venous occlusion could be distinguished. In the former the Doppler sound and instantaneous oscillation ceased. In contrast, venous occlusion produced a “damped” staccato sound and a reduced, but persisting instantaneous flow oscillation. In the free flaps, superficial epigastric artery spasm, with absent flap circulation, was detected by ultrasound Doppler flowmetry but not by the double forceps or raise tests. Although this spasm
123
MONITORING
may resolve, flap survival could not be relied upon. This was dramatically illustrated in one instance where spasm was found when the flap was raised after 1 hour but when the vessels were examined at the end of the experiment (2 hours later) the proximal superficial epigastric artery, which was not in spasm, had become occluded by clot although the anastomoses remained patent. A11 the successful flaps demonstrated the same pattern of mean arterial blood velocity and magnitude of instantaneous oscillation increasing until a steady level was reached. Late occlusion did not occur but early arterial obstruction was clearly indicated. In two flaps, in an attempt to mimic late failure, a solution of adrenaline (1:200 000) was irrigated beneath the flap after 3 hours monitoring. Constriction of small vessels produced a rapid fall in mean flow velocity to 0 on both occasions. Longer flap monitoring, after the animals had been recovered from anaesthetic, would have been desirable but was impractical in this model. Clinical evaluation is in progress and will be the subject of a later report. Conclusion
Pulsed ultrasound Doppler flowmetry with an implanted probe was valuable as a per-operative indicator of anastomotic status and showed great potential as a continuous postoperative monitor of vascularised free tissue transfers in an experimental model. Acknowledgement The ultrasound Doppler caid Limited. Chichester.
flowmeter
was kindly loaned
by Soni-
References Acland, R. D. (1972). Signs of patency in small vessel anastomosis. Surgery. 72.744. Anderson, M. F. (1981). Pulsed Doppler ultrasomc flowmeter: Application to the study of hepatic blood flow. In: Granger, D. N.. and Bulklev. G. B. (Eds) Measurement nfEIood FIoM’. Applications io the Splanchnic kirculation. p. 345.Baltimore. Williams and Wilkins. Gilsbach, J. M. (1983). lntraoperative Doppler Sonography in Neurosurger?. Vienna, Springer-Verlag. Goodstein, W. A. and Buncke, H. J. Jr. (1979). Patterns of vascular anastomosis vs success of free groin flap transfers. Plastic and Reconstructive Surgery, 64, 37. Hayhurst, J. W. and O’Brien, B.Mc. (1975). An experimental study of microvascular technique, patency rates and related factors. British Journal ofplastic Surgery, 28, 128. Jones, B. M. and Greenhalgh, R. M. (1983). The use of the ultrasound Doppler flowmeter in reconstructive microvascular surgery. Britkh Journal of Plastic Surgery, 36,245.
124 Jones, B. M. (1984). Monitors for the cutaneous microcirculation. Plastic and Reconstructive Surgery, 73,843. Jones, B. M. (1985). Predicting the fate of free tissue transfers. Annals of the Royal College of Surgeons of England, 67,63. Karkowski, J. and Buncke, H. J. Jr. (1975). A simplified technique for free transfer of groin flaps by use of a Doppler probe. Plastic and Reconstructive Surgery, 55,682. Keitzer, W. F., Licht, E. L., Brossart, F. A. and De Weese, M. S. (1972). Use of the Doppler ultrasonic flowmeter during arterial vascular surgery. &chives ofSurgery, 105,308. _ Moulder. P. V.. Teaew. _, M. J.., Manuele. V. J.., Brunswick. R. A. and Daicoff, G. R. (1977). Intraoperative Doppler coronary artery finder. Annals of Thoracic Surgery, 24.430. Shaw, W. (1981). A general survey of free flaps. Unpublished work presented at “Clinical Frontiers in Reconstructive Microsurgery”, 26th June, 1981,
BRITISH
JOURNAL
OF PLASTIC
SURGERY
The Authors J. 0. Roberts, MA, FRCS, Lecturer in Surgery, - _ Charing Cross and Westminster Medical School, London, B. M. Jones. MS. FRCS. Consultant Plastic Surgeon. The Hospital for Sick dhildren, Great Ormond Street&d University College Hospital, London. R. M. Greenhalgh, MA, MD, MChir, FRCS, Professor of Surgery, Charing Cross and Westminster Medical School, London Requests for reprints to: J. 0. Roberts, ment of Surgery, Charing Cross Hospital, London W6 8RF.
MA, FRCS, DepartFulham Palace Road.