Brirr,\A Jourd ~~/‘P/a.s~k Surgeq (I 985) 38. 369-374 ti? 1985 The Trustees of British Association of Plastic Surgeons

Peripheral neovascularisation of muscle and musculocutaneous flaps

P. G. MILLICAN and M. D. POOLE

Department of Plastic Surgery, Radcliffe Infirmary, Oxford, and the University of Oxford

Summary-The collateral vascularisation of muscle flaps covered by a skin graft and of musculocutaneous island flaps was studied in a pig model using the latissimus dorsi muscle. Vascular ingrowth from the underlying bed was stopped by using silastic sheeting. The time taken for peripheral vascularisation of flaps to render them independent of pedicle blood supply was studied. In addition microangiograms were performed at the time of sacrifice of the animals (4 weeks postop.) to give a visual conception of the sites and sizes of new vessel growth at the periphery of the flaps. Results showed significant delay, in muscle flaps with skin grafts reset onto a poor bed, in obtaining a blood supply independent of the pedicle supply.

The timing of division of a flap pedicle such that the flap will survive without the original vascular supply in the pedicle is known for most types of skin flaps. Black et al., 1978, Tsur et al., 1980 and Coleman, 1982 have all shown musculocutaneous flaps can survive without their initial axial blood supply after 7 to 10 days, if reset into an ideal bed.

The rate, quantity and quality of peripheral neovascularisation of muscle flaps covered by skin grafts in comparison to musculocutaneous flaps is not known, particularly when such flaps are trans- posed to poorly vascularised and scarred tissue.

This study was designed to attempt to elucidate the difference, if any, in these aspects of vascularisa- tion between muscle flaps covered by skin grafts and musculocutaneous flaps.

Materials and methods

The pig was chosen as the laboratory animal, since the panniculus carnosus does not contribute to the blood supply of the skin (Daniel, 1973) and so a reasonably analogous situation to the human can be observed. Each pig had a control musculocuta- neous flap raised on one side, and an equivalent muscle flap raised on the contralateral side. The latissimus dorsi flap was used, allowing isolation of the thoracodorsal artery and vein. This leash could be occluded temporarily at any time in the postoperative period without interfering with the flap edges (Millican and Poole, 1985).

Six Great White English pigs (9 weeks old, 27 kg) were used, each rooming in 2 weeks before the first

procedure. The pigs were anaesthetised with halo- thane, nitrous oxide and oxygen through an orotracheal tube.

The cutaneous element of the latissimus dorsi flap was raised as a rhomboid with 8 cm sides. The muscle was divided at the caudal border, reflected in a cephalad direction, and mobilised so that its only attachments were the avascular tendinous in- sertion into the humerus and the thoracodorsal ves- sels and nerve. Any surrounding vascular connec- tions were severed. A silastic catheter was looped over the proximal thoracodorsal leash and brought out through a separate axillary stab incision. To simulate a scarred, avascular bed a sheet of silastic (2 mm thickness) was cut to size and placed on the bed beneath the flap but not occluding the edges, so “peripheral” neovascularisation could be studied in isolation. The latissimus dorsi flap was sutured back in position with interrupted chromic gut sutures and the skin was closed with continuous nylon.

An identical procedure was performed on the contralateral side, except that the rhomboid of skin, fat, and panniculus was removed. The latis- simus dorsi muscle was covered with a split-skin graft, removed from the back of the pig more dorsally. The graft was secured with continuous nylon sutures and tieovers over a thick sponge bolus. At the end of the procedure, complete vascular isolation of the myocutaneous flap was shown by lack of filling with dye, during a 5 minute period of traction on the silastic catheters following intravenous injection (into the pig’s ear) of 10ml

369

370

of Disulphine Blue (6.2% w/v). The catheter was then released and the flap always subsequently filled with dye.

At weekly intervals for 4 weeks following eleva- tion of the flaps, each pig was reanaesthetised and given an intravenous injection of 10 ml Disulphine Blue while both thoracodorsal vessels were occluded by traction on the silastic catheters. Appearance of dye in the flap could then only be a result of neovascularisation across the wound margins. A record was made of the time taken for the flap to perfuse with dye and the direction of origin of fill. Vascularisation, independent of the thoracodorsal vessels, was assessed by giving a score to the rate of filling by dye as per Table 1.

At the end of 4 weeks, the thoracodorsal vessels were clamped (via the axillary incisions), the pig heparinised (25,000 units), the femoral artery can- nulated, and the pig perfused with a heated mixture of Micropaque and gelatin (Myers, 1975). The animal was sacrificed, cooled, and both flaps together with a surrounding 5 cm wide strip of tissue were removed by sharp dissection to the depth of the intercostal muscles. Contact X-rays were taken. Each flap was cut transversely into 1 cm slices (Fig. 1) to give at least 5 pieces of tissue, and the cross sections of the flap and its periphery

BRITISH JOURNAL OF PLASTIC SURGERY

Table 1

1. No fill. 2. Partial fill-less than 10% in 3 minutes. 3. Partial fill-greater than 10% in 3 minutes. 4. Partial fill-greater than 50% in 1 minute. 5. Total fill in 3 minutes. 6. Total fill in 1 minute. 7. Immediate fill.

subsequently X-rayed. Microangiograms were therefore obtained of transverse sections of the flap, the flap-wound interface, and tissue cephalad and caudal to the flap.

Results

All flaps remained viable, and all grafts had maximally taken at one week. There was partial loss (less than 30%) of the skin graft of several pigs at 2 weeks. Healing was complete at 3 weeks. The silastic sheeting caused seromas in 2 pigs requiring aspiration, and was partially extruded in 2 further pigs (week 3 and 4). The silastic caused a significant fibroblastic response, with a minimum of new vessel ingrowth as seen on the microangiograms.

Fig. 1

Figure l-Diagram of the flap, the surface markings and the slicing planes for microangiograms

PERIPHERAL NEOVASCULARISATION OF MUSCLE AND MUSCULOCUTANEOUS FLAPS 371

Dye studies Table 2

The results of the dye perfusion studies are listed in Tables 2 and 3. By comparison to musculocuta- neous flaps there is a delay in the development of a collateral circulation to muscle flaps covered by a skin graft. Using the Sign test, the difference between the two groups at two weeks is significant (p < 0.03).

Musculocutaneous f Pig Week I Week 2 Week 3 Week 4

At the initial surgery, dye was seen to perfuse from proximal to distal along the musculocuta- neous flap when traction on the axial vessels was released. The origin of dye fill in subsequent weeks as collateral circulation developed was from all wound edges, most often the distal.

I. 4 6 6 7 2. 1 7 7 3. 3

z 7 7

4. 3 6 7 7 5. 5 6 I 7 6. 6 6 7 7

Maple jap and skin graft Pig Week I Week 2 Week 3 Week 4

I. 3 4 7 7 2. 1 2 6 I 3. 2 4 6 7 4. 1 3 7 I 5. 1 2 4 6 6. i 2 3 6

Microangiography

The major large arteriolar growth supplying col- lateral circulation was at the muscle/muscle inter- face. In the musculocutaneous flaps, new vessel formation also occurred at the dermal, subdermal and to a much lesser extent subcutaneous level. However at 4 weeks post-surgery, when the animal was sacrificed, these vessels were finer than at the muscle interface of the wound. Visually, this sug- gested that neovascularisation from the surround- ing skin and subdermal tissues supplied less col- lateral flow than that from surrounding muscle (Fig. 2).

By contrast, the latissimus dorsi muscle flap and skin graft had an extremely minor subdermal circulation with little vascularisation from the

Table 3

Musculocu taneous ftaps

t A11 filied by 2 weeks. + 50% filled at 1 week. * 2 cases without silastic-immediate fill at 1 week.

Muscle jap and skin graft

+ All filled at 4 weeks. * 66% filled at 3 weeks. * None had total fill at 2 weeks.

Fig. 2

Figure 2---Neovascularisation at the flap edges showing this is mostly at the subdermal and muscle interfaces.

372

surrounding dermis and subcutaneous tissue. Major collaterals were present in the muscle/ muscle edges. At 4 weeks the flap muscle had “undermined” the circumferential skin edges (by contraction of the surface skin graft), to develop an interface (but not a vascular one) with fat also (Fig. 3).

In 2 further pigs, latissimus dorsi musculocuta- neous flaps, as described above, replaced onto beds without silastic gave immediate fill of dye at one week. (Grading 7, Table 1). Microangiograms showed the base to have a leash of new vessel anastomoses which communicated widely with the overlying muscle, having an extremely large surface area for circulation exchange, as compared with the lateral edges (Fig. 4).

Discussion

This initial study, though hampered by lack of numbers, does show some interesting points. Our studies show peripheral vascularisation of muscle flaps to take significantly longer to develop than musculocutaneous flaps. Collateral circulation does become sufficient at four weeks, however, to

BRITISH JOURNAL OF PLASTIC SURGERY

fill immediately the flap and skin graft with dye (with pedicle vessel occlusion). The major site of the new anastomoses on angiography is muscle to muscle. Superiority of musculocutaneous flap neo- vascularisation must be attributed to the extra layer of dermal and subdermal tissue with their associated vascular plexus.

Previous studies on muscle or musculocutaneous flaps have inset them into an ideal base, and given varying quantitative results of the time taken for flaps to survive completely. Coleman (1982) de- scribed musculocutaneous flaps as being indepen- dent of their axial supply in a minimum of 6 days, if placed on an ideal bed. Black et al. (1978) reported full survival of 5 cm square superior epigastric flaps without pedicle at 8 days with an ideal bed. Clinically of course one is often dealing with less than ideal beds.

One cannot draw conclusions about “viability” from our dye and angiographic studies as dye infusion with the pedicle vessels occluded does not give a quantitative result of flap viability, but an indication of the rate of collateral perfusion. Total perfusion must be equivalent to a viable flap after occlusion of the main vessel but extrapolation of lesser perfusion to flap viability is not valid.

Fig. 3

Figure 3-A muscle flap and skin graft showing the poor neovascularisation in the superficial layers and most of it being at the muscle interface.

PERIPHERAL NEOVASCULARISATION OF MUSCLE AND MUSCULOCUTANEOUS FLAPS 373

Clinical relevance Khoo and Bailey (1982) reported a number of clinical muscle containing flaps which underwent partial necrosis when the axial vessel thrombosed more than 7 days post surgery. All had surviving muscle over the bed of the flap. They suggested that the bed was highly significant in development of collateral circulation, and peripheral healing may not be sufficient to maintain flap viability.

We believe, similarly, that if the bed on which the flap lies is poorly vascularised (due to large areas of scar, cortical bone, irradiated tissue, etc.) then the peripheral inset is of importance in flap vascularisa- tion. Indeed, we have seen one unusual case of total loss of the distal end of a grafted gastroc- nemius muscle flap which was used to cover cortical bone and inset into scarred tissue, when the muscle belly (pedicle) was divided for an ortho- paedic procedure over 18 months after the flap was done. Recently Fisher and Wood (1984) have des- cribed late necrosis of the latissimus dorsi muscle flap covered by a skin graft on the leg, which sounds very similar to our own case.

Many clinical situations require muscle- containing flaps to cover compound avascular surfaces not suitable for direct grafting. Our find- ings would suggest that when using muscle flaps it

is important where possible to have flap muscle abutting healthy muscle, rather than scarred skin and fat, particularly if further surgery is likely which will interrupt the pedicle blood supply. Collateral circulation to a muscle flap plus skin graft takes longer to develop than to a musculo- cutaneous flap, but should eventually be sufficient to maintain the flap independent of axial supply. At least 4 weeks should be left before reexplora- tion, leaving incision of wound edges to a minimum.

Questions which may be important clinically but which were not answered by this study are (i) the state of the microcirculation after complete wound healing, scar maturity, muscle atrophy and con- tinuing axial supply; (ii) the ratio of dermal to muscular collateral blood flow; and (iii) viability of very large muscle flaps after delayed loss of blood supply.

Acknowledgements

The advice and help of Dr G. M. Ardran and Mrs W. Hills (Dept. of Radiology-University of Oxford) and of Dr C. Young, Dr J. Hopewell and staff of the Research Institute, Churchill Hospital. Oxford is gratefully acknowledged.

Fig. 4

Figure 4-A musculocutaneous flap without silastic covering the “bed”extensive neovascularisation occurs at the deep surface of the flap.

374 BRITISH JOURNAL OF PLASTIC SURGERY

References

Black, M. J., Chait, L., O’Brien, B., Sykes, P. and Sharzo, L. (1978). How soon may the axial vessels of a surviving free flap be safely ligated: a study in pigs. British Journal of Plastic Surgery, 31, 295.

Coleman, J. J. III. (1982). Long term evaluation of muscle and musculocutaneous flaps. In Mathes, S. J. and Nahai, F. (1982). Clinical Applications for Muscle and Musculocuta- neous Flaps. 1st edition. St. Louis, C. V. Mosby.

Daniel. R. K. (1973). The free transfer of skin flans bv ~ I _ _

microvascular anastomoses. Plastic and Reconsfructive Sur- gery, 52, 16.

Fisher, J. and Wood, M. D. (1984). Late necrosis of a latissimus dorsi free flap. Plastic and Reconstructive Surgery, 74, 274.

Khoo, C. T. K. and Bailey, B. N. (1982). The behaviour of free muscle and musculocutaneous flaps after early loss of axial blood supply. British Journal of PIustic Surgery, 35, 43.

Millican, P. G. and Poole, M. D. (1985). A pig model for investigation of muscle and myocutaneous flaps. British Journal of Plastic Surgery, 38,364.

Tsur, H., Daniller, A. and Strauch, B. (1980). Neovascularisa- tion of skin flaps: Route and timing. Plastic and Reconstruc- tive Surgery, 66, 85.

The Authors

P. G. Millican, FRCS, FRACS, Overseas Fellow, Department of Plastic Surgery, Radcliffe Infirmary, Oxford.

M. D. Poole, FRCS, FRACS, Consultant, Department of Plastic Surgery, Radcliffe Infirmary, Oxford.

Requests for reprints to: Mr M. D. Poole, Plastic Surgery Department, Radcliffe Infirmary, Oxford, U.K., OX2 6HE.

This work was supported by a grant from the Medical Research Fund, University of Oxford.