MAXIMIZING FLAP SURVIVAL IN A PREFABRICATION MODEL USING EXOGENOUS AND ENDOGENOUS BFGF: A NEW APPROACH

FREDERICK J. DUFFY, at., M.D., PETER K.M. MAITZ, M.D., CHARLES A. HERGRUETER, M.D., and JULIAN J. PRIBAZ, M.D.

Flap prefabrication is dependent on the neovascular re- sponse that occurs between the implanted arteriovenous pedicle and the recipient tissue. Augmentation of this neo- vascular response with angiogenic growth factors would maximize flap survival and minimize the interval between pedicle implantation and flap rotation. Maximizing the bio- logic activity of endogenous growth factors would likewise positively impact upon flap survival. This study examined the role of basic fibroblast growth factor, a known potent angio- genic growth factor, on flap survival in a rabbit ear prefabri- cation model. Sucrose octasulfate, a substance that binds

basic fibroblast growth factor, stabilizes it, and protects it from degradation, was also studied to determine its impact on flap survival. Flap survival was increased using basic fi- broblast growth factor, sucrose octasulfate, and the two sub- stances combined together. The use of substrates designed to maximize the biologic activity of endogenous growth fac- tors, rather than relying on the artificial addition of exogenous growth factors, represents a new approach in the search for methods that will improve flap survival.

0 1997 Wiley-Liss, Inc. MICROSURGERY 17:176-179 1996

T h e need for thin, vascularized flaps far exceeds the sup- ply, especially in reconstructive situations involving the head and neck and upper extremity. Flap prefabrication would address this shortage, allowing for the creation of customized, thin, vascularized flaps by implantation of a vascular pedicle into a predetermined donor territory. The concept of flap prefabrication is supported by numerous animal models as well as a number of anecdotal clinical reports. We have developed our own flap prefabrication model using an exteriorized vascular pedicle based on the central arteriovenous bundle of the rabbit ear that was de- signed to minimize both donor deformity and pedicle trauma.'

Flap prefabrication is dependent upon the neovascular response that occurs after a vascular pedicle has been im- planted into a recipient tissue. Maximizing this neovascular response would aHow for improved flap survival while di- minishing the interval between pedicle implantation and flap rotation. We would predict that the addition of an an- giogenic growth factor, such as the potent angiogenic pep- tide, basic fibroblast growth factor (bFGF), would increase the neovascular response following pedicle implantation. At least one existing study examines the impact of bFGF on

From the BrighamiChildren'dHaward Division of Plastic Surgery

*Correspondence to: .Julian J. Pribaz, M.D., Division of Plastic Surgery, Brigham and Women's Hospital, 75 Francis Street, Boston, MA 021 15-61 95.

Received for publication September 6, 1995; revision accepted April 6, 1996.

random skin flap survival.' The role of this angiogenic growth factor in a prefabrication model remains unknown, however.

If it were possible to maximize the biologic activity of endogenous growth factors, flap viability would almost cer- tainly be positively impacted upon. An intriguing group of experiments was reported in a recent paper that questioned the pathophysiology of duodenal ulcers and the mechanism by which they heal.' The studies provided support for the probable mechanism of sucralfate, a widely used anti-ulcer drug. The sucralfate was found to have an extremely high affinity for bFGF, binding to it and protecting it from deg- radation. Folkman's group postulated that the drug may work by binding endogenous bFGF, concentrating it in the ulcer bed, and maximizing the peptide's angiogenic and wound-healing potentials. Based on these data, we hypoth- esize that sucrose octasulfate, the soluble counterpart of sucralfate, will stabilize endogenous bFGF in our flap pre- fabrication model, improve the neovascular response, and ultimately improve flap survival.

MATERIALS AND METHODS Eighteen female New Zealand white rabbits weighing

2.5 to 3.0 kg were used in this study. Anesthesia was in- duced by intramuscular Ketamine and Xylazine. All ani- mals received 150,000 units of intramuscular penicillin be- fore each procedure. The rabbits were housed in single cages and received normal diets postoperatively. All pro- cedures were done under sterile conditions.

0 1997 Wiley-Liss, Inc

Growth Factors in Prefabricated Flaps 177

1 - J

Figure 7. First stage: A tubed, exteriorised arteriovenous pedicle is implanted into the rabbit's neck. The exposed 7 cm pedicle is sup- ported by a 1 x 4 cm strip of auricular cartilage and a 1 x 4 cm gelfoam strip (carrier for the various agents) is implanted between the pedicle and skin.

First Stage: Creating an Exteriorised Arteriovenous Pedicle

An arteriovenous pedicle was created from the central auricular artery and vein of each ear and then tubed and exteriorised in a surrounding skin tube as detailed in our model previously described. The pedicle remains attached at the base of the ear and the distal exposed 1 cm of arte- riovenous bundle, supported by a 1 X 4 cm cartilage tem- plate, is implanted into a subcutaneous pocket on the rab- bit's neck (Fig. 1).

At this point, the animals were divided into six exper- imental groups of three rabbits (six flaps) each as outlined below:

(NS = normal saline; SOS = sucrose octasulfate) Group 1 (control 1): pedicle implantation only. Group 2 (control 2): pedicle implantation plus a 1 x 4

cm piece of plain gelfoam. A similar-sized piece of gelfoam will be used in the remaining groups as a carrier for the various agents being tested.

Group 3: bFGF-soaked gelfoam (0.1 mcg bFGF/ml NS) .

Group 4: bFGF-soaked gelfoam ( 1 .O mcg bFGF/ml NS) .

Group 5: sucrose octasulfate-soaked gelfoam (100 mcg SOS/ml NS).

Figure 2. Second stage: Transfer of prefabricated 1 x 6 cm flap (including cartilage and remains of gelfoam) to the ipsilateral ear at two weeks.

Group 6: sucrose octasulfate and bFGF-soaked gelfoam (100 mcg SOSlml NS and 1 .O mcg bFGF/ml NS).

The two dosages of bFGF were chosen within the range of prior dosages that have been shown to be effective in in vivo and in vitro

Second Stage: Transfer of Prefabricated Neck Flap to Ear

After a period of two weeks had passed, a 1 x 6 cm flap was raised based on the pedicle and rotated to the ipsilateral ear as outlined in our model.7 (Fig. 2 ) . This transfer includes what is left of the underlying gelfoam.

Third Stage: Assessment of Flap Survival Seven days following flap transfer, the percentage of

flap survival was assessed by tracing the total flap size and the portion of the flap that survived on Esmark bandage of uniform thickness and under no tension and then calculating the percentage of flap survival.

178 Duffy et al.

TABLE 1. Flap Survival

Group

1. Control I 2. Control II (gelfoam only) 3. bFGF (0.1 mcgiml) 4. bFGF (1 .O mcg/ml) 5. S.O.S. 6. S.O.S.

P Value Standards Standard % Flap survival Mean % deviation error vs. Group I vs. Group 2

31,35,67,44,42,47 44.3 12.5 5.1 43,55,55,44,29,34 43.3 10.6 4.3 91,48,69,66,48,47 61.5 17.4 7.1 0.079 0.054 91,76,32,67,68,71 67.5 19.5 7.9 0.034 0.024 63,69,59,72,74,82 69.8 8.2 3.3 0.002 <0.001 68,56,65,79,75,81 70.6 9.5 3.9 0.002 <0.001

RESULTS All animals survived the protocol. No flap hematomas

or infections occurred. In addition, none of the pedicles were mutilated or disturbed by the rabbits. The percentage flap survival and statistical analysis for the various groups are listed in Table 1. Statistical significance was assessed using a t-test and comparing the experimental groups to both control groups (groups 1 and 2). Statistically improved survival was demonstrated in group 4 (the higher concen- tration of bFGF), group 5 (the sucrose octasulfate group), and group 6 (sucrose octasulfate in combination with bFGF). Maximal flap survival was demonstrated in the flaps treated with both sucrose octasulfate and bFGF.

DISCUSSION Flap prefabrication is an interesting concept that is sup-

ported by a number of experimental animal models as well as a small sampling of anecdotal clinical cases. It would have potential utilization in clinical scenarios that require a thin, vascularized flap that wasn’t readily available. These situations usually involve head and neck and upper extrem- ity reconstruction. The technique is likely to become an acceptable reconstructive option only if a prefabricated flap would be of sufficient size to adequately deal with a larger defect requiring a thin flap. Ideally, the interval between pedicle implantation and flap rotation should be sufficiently short to allow for staged reconstructions to proceed at a reasonable pace. Flap prefabrication is dependent on the neovascular response that occurs following pedicle implan- tation. Maximizing this neovascular response would in- crease flap survival and ultimately increase the likelihood that this technique might serve as a viable, clinical recon- structive option.

In the past,7 we have shown that essentially 100% flap survivai is possible using this model if the interval between pedicle implantation and flap transfer is at least 6 weeks. Shortening this interval in a reliable way could certainly facilitate reconstructive efforts, however. We have demon- strated the beneficial effect of delay incisions in enhance- ment of prefabricated flap survival and found that the tim- ing of the delay incisions were important, with improved flap survival if the delays were performed one week before or one week after pedicle implantation.’” In this group of

experiments we have attempted to study the effects of an angiogenic growth factor in an effort to shorten the interval between pedicle implantation and flap transfer. Our results demonstrate a statistically significant improvement in flap survival when the higher concentration of the angiogenic peptide bFGF was used, when sucrose octasulfate was used, and when the two were used together. Presumably, this improved flap survival is a result of more effective or more abundant angiogenesis mediated or stimulated by the bFGF.

Many authors have studied basic fibroblast growth fac- tor in recent years in a variety of different models.”-‘4 Hayward et al.” were unable to document any increased angiogenesis in flaps treated with bFGF, while Kim et al.” documented increased angiogenesis in bFGF-treated tis- sues, especially when ischemic. The exact mechanism by which bFGF increases angiogenesis is unclear, though it is known to stimulate endothelial cell mitogenicity and chemotaxis. l6 It is clear, however, that the wound healing cascade (including angiogenesis) is a very complicated phe- nomenon based on the interrelationship of a variety of growth factors as well as extracellular matrirc proteins.I7 Adding large, exogenous amounts of bFGF to a healing wound at a fixed point in the wound healing cascade is a very simplistic approach. In addition, utilization of exoge- nous growth factors could have untoward short-term or long-term effects that have not yet been identified.

It is interesting to compare the results of this study using angiogenic factors in flap prefabrication with our previously published work on the effects of the delay phenomenon in the same model.’” In her invited discussion on a paper by Uhl et al. ,I4 Kerrigan” summarizes the published work comparing the effects of bFGF and the delay phenomenon on flap survival. In most random flaps models, delay was a more effective tool for improving flap viability. However, in our rabbit ear prefabrication model, we have observed that the mean flap survival in our delay experiments (66.3% and 63.17%) was very similar to the improved survival seen in groups 4 through 6 in the present experiment (67.5%, 69.8%, and 70.6%). As Kerrigan points out in her discus- sion, the mechanism of increased flap survival in either group remains unclear but is worthy of further study.

Augmentation of the biologic activity of endogenous growth factors would seem to be a more “homeopathic”

Growth Factors in Prefabricated Flaps 179

approach. Because of Folkman’s’ work with sucralfate, this may prove to be possible with bFGF. Basic fibroblast growth factor is known to have a high affinity for heparin. l9

It is theorized that bFGF is bound to heparin proteoglycans in the extracellular matrix and is released following an in- jury stimulus.*’ Interestingly, sucrose octasulfate is very similar structurally to heparin and also binds bFGF and stabilizes it. It is possible that the improved flap survival seen in the sucrose octasulfate groups in this protocol is a result of this localization and stabilization of bFGF which allows for its maximal stimulation of the angiogenic re- sponse and therefore improved flap survival.

CONCLUSION The use of substrates designed to maximize the biolog-

ical activity of endogenous growth factors represents a new and potentially valuable methods of manipulating the growth factor milieu. It negates the potential adverse con- sequences of exogenously applied growth factors and in- stead strives to maximize the effect of the host’s endoge- nous growth factors. Further experimental trials and eventual clinical trials are indicated.

REFERENCES 1. Erol 00: The transformation of a free skin graft into a vascularized

pedicle flap. Plast Reconstr Surg 58:470-477, 1976. 2. Shen ZY: Greater omentum-cutaneous axial flap: A method to create

transferable skin flap. Chin Med J 94:718, 1981. 3. Shen TY: Vascular implantation into skin flaps: Experimental study

and clinical application: A preliminary report. Br. J Plast Surg 40: 403-410, 1987.

4. Morrison WA, Dvir E, Doi K, Hurley JV, Hickey MJ, O’Brien B: Prefabrication of thick transferable axial-pattern skin flaps: An exper- imental study in rabbits. Br J Plast Surg 44593-598, 1991.

5 . Hyakusuoku H, Okubo M, Umeda T, Fumiiri M: A prefabricated hairbearing island flap for lip reconstruction. Br J Plast Surg 40:37- 39, 1987.

6. Khouri RK, Upton J. Shaw WW: Prefabrication of composite free

flaps through staged microvascular transfer: An experimental and clin- ical study. Plast Reconsfr Surg 87:108-115, 1991.

7. Duffy FJ, Pribaz JJ, Hergrueter CA, Maitz PKM: Flap prefabrication using an exteriorised vascular pedicle in a rabbit ear model. Br J Plust Surg 46:279-284, 1993.

8. Khouri RK, Brown DM, Leal-Khouri SM, Tark IKC, Shaw WW: The effect of basic fibroblast growth factor on the neovascularisation pro- cess: Skin flap survival and staged flap transfers. Br J Plast Surg 44585-588, 1991.

9. Folkman J, Szabo S, Stovroff M, McNeil P, Li W, Shin Y: Duodenal ulcer. Discovery of a new mechanism and development of angiogenic therapy that accelerates healing. Ann Surg 214:414-427, 1991.

10. Maitz PKM, Pribaz JJ, Duffy FJ, Hergrueter CA: The value or the delay phenomenon in flap prefabrication: An experimental study in rabbits. Br J Plast Surg 47:149-154, 1994.

11. Hayward PG, Alison WE, Carp SS, Hui PS, Robson MC: Local infiltration of an angiogenic growth factor does not stimulate the delay phenomenon. Br J Plast Surg 44526-529, 1991.

12. Im MJ, Kim Ys, Edwards RJ, Hoopes JE, Fenselau A: The effect of bovine basic fibroblast growth factor on skin flap survival in rats. Ann Plust Surg 28:242-245, 1992.

13. Ishiguro N, Yabe Y, Shimizu T, Iwata H, Miura Y: Basic fibroblast growth factor has a beneficial effect on the viability of random skin flaps in rats. Ann Plast Surg 32:356-360, 1994.

14. Uhl E, Barker J, Bondar I, Galla T, Lehr H, Messmer K: Improve- ment of skin flap perfusion by subdermal injection of recombinant human basic fibroblast growth factor. Ann Plust Surg 32361-366, 1994.

15. Kim L, Ogden L, Barket J , Anderson G, White S, Uhl E, Kamler M, Tobin G . Basic fibroblast growth factor induces angiogenesis in isch- emic tissues. Surg Forum 77:632-635, 1991.

16. Gross JL, Moscatelli D, Rifkin DB: Increased capillary endothelial cell protease activity in response to angiogenic stimuli in vitro. Proc Nut1 Acad Sci USA 80:2623-2627, 1983.

17. Ingber DE, Madri JA, Folkman J: Endothelial growth factors and extracellular matrix regulate DNA synthesis through modulation of cell and nuclear expansion. In Vitro Cell Dev Biol23:387-394, 1987.

18. Kenigan CL: Invited discussion, improvement of skin flap perfusion by subdermal injection of recombinant human basic fibroblast growth factor. Ann Plast Surg 32:365-366, 1994.

19. Shing Y, Folkman J, Sullivan R, Butterfield C, Murray J, Klagsburn M: Heparin affinity: Purification of a tumor-derived capillary endo- thelia cell growth factor. Science 223:1296-1298, 1984.

20. Baird A, Ling N: Fibroblast growth factors are present in the extra- cellular matrix produced by endothelia cells in vitro: Implications for a role of heparinase-like enzymes in the neovascular response. Bio- chem Biophys Res Commun 42:428-435, 1987.