Veterinary Surgery, 20, 6, 379-384, 1991
Abdominal Wall Reconstruction with a Vascular External Abdominal Oblique Myofascial Flap LISA G. ALEXANDER,
MICHAEL M. PAVLETIC,
DVM, Diplomate ACVS,
and STEPHEN J. ENGLER, VMD
A myofascial island flap for abdominal wall reconstruction was based on the lumbar component of the external abdominal oblique muscle and supplied by a major neurovascular pedicle consisting of branches of the cranial abdominal artery, cranial hypogastric nerve, and a satellite vein. The flap was elevated and sutured into a 10 cm X 10 cm body wall defect in five dogs. The dogs were observed for 26 to 28 days. Abdominal wall contour and function were preserved. All dogs developed seromas, two of which became infected. One dog developed a hernia at the dorsal margin of the flap, which was repaired. At necropsy, there was no evidence of dehiscence in any of the dogs. Loose adhesions of omentum to the inner surface of the flap occurred in four dogs. Results of histologic examination confirmed the clinical impression of flap viability. The myofascial island flap has a wide range of mobility over the ventral and caudal areas of the abdomen and lateral thoracic wall. It has potential clinical use for reconstruction of defects within its arc of rotation.
wall defects occur as a result of trauma, infection, “en bloc” excision of tumor masses, or herniation with associated muscular retraction and atrophy.’-8 Such defects may be impossible to close primarily without creating secondary defects or incisional tension. An ideal reconstructive technique would provide strong functional support for the abdominal contents under minimal tension, with the lowest associated morbidity and mortality.? Previously reported methods used exogenous synthetic or biologic materials for reconstruction. Synthetic prostheses such as stainless steel, polyester, polyethylene, nylon, microporous polytetrafluoroethylene, and carbon-polycaprolactone composites have been used in animals with varying success.’ 3-5 ‘ - I 3 In humans, fascia1 grafts, dermal grafts, myofascial flaps, and myocutaneous flaps have been used for abdominal wall reconstruction.’ I4-l7Cutaneous, muscular, and myocutaneous flaps have been evaluated to restore large defects of the trunk and extremities and to cover exposed bone in We found no published reports in which muscular flaps were specifically evaluated for repair of large abdominal wall defects in dogs. The purpose of this study is to identify and evaluate a vascular myofascial island flap based on the lumbar component
of the external abdominal oblique muscle for use in fullthickness abdominal wall reconstruction. Materials and Methods A nat omic Considerat ions
The external abdominal oblique muscle is long, broad, and flat. with fibers running caudoventrally. It consists of costal and lumbar components. The costal component originates segmentally from the fourth or fifth rib through the thirteenth rib. The lumbar component originates in the thoracolumbar fascia along the iliocostalis muscle. Ventrally and caudally, the aponeuroses of the two components contribute to the external rectus fascia, the external inguinal ring, and the prepubic tendon. The cranial branch of the cranial abdominal artery supplies the middle zone of the lateral abdominal wall and is accompanied by the cranial hypogastric nerve and a satellite vein. The deep branch of the deep circumflex femoral artery anastomoses with the cranial and caudal abdominal arteries and is the main supply to the caudodorsal fourth of the abdominal wall. It is accompanied by a satellite vein and is joined by the lateral cutaneous femoral nerve.”
From the Department of Surgery, Tufts University School of Veterinary Medicine, North Grafton (Alexander, Pavletic), and the Department of Pathology, Tufts Veterinary Diagnostic Laboratory, Jamaica Plain (Engler), Massachusetts. Supported by the Tufts University School of Veterinary Medicine Faculty Fund. No reprints available.
VASCULAR MYOFASCIAL FLAP
Fig. 1 . Major steps in elevation and transfer of an external abdominal oblique myofascialisland flap. (A) Location of the lumbar component of The flap, with three edges severed, reflected to show the locationof the major neurovascular pedicle. the external abdominal oblique muscle. (8) (C) The completely elevated flap tethered by the major neurovascular pedicle. (D) The created defect in relation to the elevated flap. (E) The inner fascial surface of the flap sutured to the dorsal edge of the defect. (F) The flap completely sutured over the defect.
Anatomic Studies Five anesthetized dogs were injected intravenously (IV) with 10 mL sodium heparin (1000 U/mL). Ten minutes later, the dogs were euthanatized. Both external abdominal oblique muscles were dissected to confirm their in situ anatomy, blood supply, and innervation, and to determine the mobility of a myofascial island flap based on the lumbar portion of the muscle. The regional anatomy was examined closely to determine a surgical approach based on the dominant vascular pedicle. The arterial supply to eight of the muscles was injected with barium sulfate and the muscles were radiographed to determine the gross vascular distribution within the proposed myofascial island flap. Abdominal Wall Reconstruction Five conditioned adult mixed-breed dogs (2 males and 3 females) were premedicated with atropine (0.04 mg/kg
intramuscularly [IM]) and acetylpromazine (0.05 mg/kg IM). General anesthesia was induced with thiamylal sodium (10 mg/kg IV) and maintained with halothane and oxygen inhalation. The left abdominal wall was clipped of hair and prepared for surgery with alternating chlorhexidine and isopropyl alcohol surgical scrubs. Using aseptic technique, a 20-cm left paracostal skin incision was made from the level of the epaxial muscles to the ventral midline, 5 cm caudal to the thirteenth rib. The skin and subcutaneous tissues were retracted to expose the lumbar portion of the external abdominal oblique muscle (Fig. 1A). The fascial edges of the lumbar portion of the muscle were divided ventrally and caudally, leaving a 0.5 cm margin of fascia along the muscular edge. The muscle was undermined and the neurovascular pedicle, consisting of branches of the cranial abdominal artery and cranial hypogastric nerve and a satellite vein, was identified craniodorsally, just caudal to the thirteenth rib
ALEXANDER, PAVLETIC, AND ENGLER
suture removal on day 14. The bandages were changed daily and the drains were removed when drainage was minimal. The dogs were euthanatized with an overdose of sodium pentobarbital IV on day 26 (2 dogs), day 27 ( 2 dogs), and day 28 ( 1 dog). Necropsy examinations were performed immediately after euthanasia. The abdominal wall was examined for evidence of adhesions or dehiscence. Tissue samples were harvested from the central myofascial island flap, the cranial abdominal wall-flap interface, and the linea alba-flap interface. Specimens of the contralateral external abdominal oblique muscle were submitted as controls. The tissues were affixed to tongue depressors with pins to prevent contraction and distortion. placed in 10%buffered formalin. stained with hematoxylin-eosin. and examined histologically. Fig. 2. A partially elevated lumbar external abdominal oblique myofascial island flap, viewed from the caudal aspect, showing the location of the major neurovascular pedicle (N) in relation to the caudodorsal (CD) and caudoventral (CV) corners of the flap, and the 13th rib (R). The solid arrow points toward the dog's head, and the open arrow indicates the dorsal midline.
(Figs. 1B and Fig. 2 ) . The dorsal fascial attachment was divided, and the muscle was severed at the level of the thirteenth rib (Fig. 1C). A 10 cm X 10 cm defect was created in the left abdominal wall, extending caudally from the umbilicus and laterally from the linea alba (Fig. ID). The myofascial island flap, tethered by its neurovascular pedicle, was moved ventrally to cover the defect. The inner fascial surface of the flap was sutured to the dorsal fascial edge of the defect with simple interrupted sutures of 2-0 monofilament nylon placed 5 mm from the dorsal fascial edge and 5 mm to 7 mm apart. Care was taken to avoid the vascular branches during suture placement (Fig. 1 E). The cranial, ventral, and caudal edges of the flap were sutured to the borders of the defect in a similar fashion (Fig. 1F). One 6.35 mm (W') latex drain was anchored dorsally over the flap and exited through a stab incision 5 cm caudoventral to the skin incision. The subcutaneous tissues were closed with 3-0 nylon i n a simple continuous pattern. The skin incision was closed with 3-0 nylon in a simple interrupted pattern. A light bandage was applied to cover the incision and drain. The dogs were supervised during anesthetic recovery and returned to their runs when they were ambulatory. Butorphanol (0.2 to 0.4 mg/kg IM) was administered as needed for postoperative pain.
I n the preliminary anatomic study it was demonstrated that the lumbar component of each external abdominal oblique muscle had a reliable arterial supply originating from the cranial abdominal artery. A satellite vein and branches of the cranial hypogastric nerve accompanied the arterial supply. The major neurovascular pedicle was located at the craniodorsal edge of the muscle (Fig. 3). The vasculature arborized from this location to supply the length of the lumbar component of the external abdominal oblique muscle. Occasionally, a secondary blood supply, a branch of the deep circumflex femoral artery, existed caudodorsally as a lesser vascular pedicle. In all
Postoperutiw Cure und Evuliiution The dogs were housed individually in runs and fed a standard hospital diet. General attitude, physical examination findings, body temperature, body wall contour, and quality and quantity of drainage were recorded daily. The incision and drainage sites were kept bandaged until
Fig. 3. A radiograph made after injection of barium sulfate into the arterialside of the major neurovascular pedicle of the left lumbar component of the external abdominal oblique muscle illustrates the gross vascular distribution. A continuous line of wire suture material(arrow) marks the level of the 13th rib. Cranial (CR), caudal (CD), dorsal (D), ventral (V).
VASCULAR MYOFASCIAL FLAP
Fig. 4. Body wall contour of a dog with a left lumbar external abdominal oblique rnyofascial island flap, day 28. The incision is still visible (arrows). (A) Ventral view. (B) Lateral view.
cadaver and surgical specimens, one or two direct cutaneous arteries originated from the external abdominal oblique vasculature at approximately the midpoint of the proposed flap. In the surgical specimens, these arteries were ligated without apparent injury to the overlying skin. In the cadaver specimens, the myofascial island flap could be mobilized to reach the ninth or tenth rib cranially, past the linea alba ventrally, and to the inguinal ring caudally. The myofascial island flap was much more mobile and elastic in the surgical specimens than in the cadavers. In the surgical study, the major neurovascular pedicle did not vary from that of the cadaver specimens. The lesser vascular pedicle was present in two dogs and was ligated to facilitate mobilization. The fascia1 edges of the flaps held suture material well. One dog had herniation of small intestine into the subcutaneous space on day 3. With the dog under anesthesia, the site of herniation was identified at the dorsal margin of the original defect. The small intestine was replaced, the defect was resutured, and the dorsal border of the flap that overlapped the defect was sutured securely to the underlying internal abdominal oblique muscle with 2-0 nylon. Small-to-large serum ac-
cumulations in the left inguinal area that could be expressed easily through the surgical drain developed in four dogs. The fluid accumulations in two dogs resolved completely by day 10. In one dog, the left inguinal serum accumulation did not communicate with the surgical drain. It was managed without drainage and resolved slowly; it was small but still present at necropsy. Two dogs became febrile (40.8"C and 39.4"C) and purulent drainage developed on day 6. Cultures were obtained and trimethoprim-sulfadiazine ( I5 mg/kg orally twice daily) was administered. From one of the dogs, group G streptococci, Acenitohucter crkoaceticus, and Siaphylococcus intermediiis were cultured, all of which were sensitive to tnmethoprim-sulfadiazine. Streptococcus equi cultured from one dog was sensitive to cefadroxil. Each infection responded to additional drainage, bandaging, and appropriate antimicrobial therapy. At euthanasia, each dog had excellent body wall contour, with good muscle tone palpable in all the myofascial island flaps (Fig. 4). At necropsy, the anatomy and rnobility of the right (untreated) external abdominal oblique muscle was identical to the cadaver and surgical speci-
ALEXANDER, PAVLETIC, AND ENGLER mens. The external surfaces of the myofascial island flaps were covered with granulation tissue. Internally, the omentum was loosely adhered to the flap margins in four dogs. The remainder of the inner surfaces of the flaps was covered with a dense, smooth, white tissue. No laxity or discontinuity was detected at any flap-defect interface. Muscular tissue of the flaps and the defect edges was grossly normal. Histologic sections of the right external abdominal oblique muscles were composed of normal skeletal muscle. Sections of the left external abdominal oblique myofascial island flap showed the muscle to be covered with maturing fibrous connective tissue on the inner and outer surfaces and to be uniformly viable. Individual myofiber necrosis in sections at margins or near suture sites suggested that cell death was due to compression-induced local vascular compromise. There was a zone of proliferative, maturing fibrous connective tissue at the interface of the flap and the body wall. In two dogs, there was evidence of suppurative inflammation involving the outer fibrous covering of the flap, suggesting an external source and limitation of the infection to the subcutaneous space.
Discussion It has been stated that the ideal muscle flap should be broad, thin, reliable, and available bilaterally, and it should create minimal donor site morbidity.’ The results of this study indicate that the external abdominal oblique myofascial island flap meets these criteria. Study of the cadaver and surgical specimens demonstrated the reliable existence and location of the major neurovascular pedicle to the lumbar components of both external abdominal oblique muscles. Only the left lumbar external abdominal oblique muscle was used in the surgical study to facilitate record keeping and documentation. The right lumbar external abdominal oblique muscle served as an anatomic control and was examined at necropsy. In two animals in the surgical study, a lesser vascular pedicle supplying the caudodorsal part of the lumbar component of the external abdominal oblique muscle was sacrificed to maximize mobilization without apparent compromise of the flap (i.e., the blood supply from the major vascular pedicle alone was sufficient). The reliable existence of direct cutaneous vessels as branches of the external abdominal oblique vasculature suggested that a myocutaneous flap could be developed, based on the same dominant neurovascular pedicle as the myofascial island flap. The lumbar portion of the external abdominal oblique muscle was elastic and mobile. It was capable of extending past the defect margins in this study. This suggests that the myofascial island flap developed from it has potential clinical use for caudal thoracic wall reconstruction, caudal
abdominal wall defects, and larger abdominal wall defects past the linea alba. Additional mobility also indicates that the dorsal part of the flap may be useful for reconstruction even if the ventral margin is damaged or missing. The flap can be mobilized and rotated to cover dorsal defects, and it could be useful as a myocutaneous composite graft. A lumbar external abdominal oblique myofascial island flap could be used to reinforce weak areas of the abdominal wall that are not full-thickness defects. There also is potential for bilateral elevation and transfer of flaps for repair of massive ventral abdominal wall defects. The procedure for harvesting the flap was easily performed without specialized instrumentation. The cranial, caudal, and ventral borders of the flap held sutures well. The dorsal margin of the defect overlapped by the flap appeared most susceptible to dehiscence because fullthickness sutures were not used to anchor the myofascial island flap out of concern for interruption of the vascular supply. The dog that developed a herniation in this location had episodes of “explosive” activity in the immediate postoperative period despite confinement, thus placing the reconstruction site under stress. Dead space was created during elevation and transfer of the flap, predisposing it to seroma formation.23Postoperative fluid accumulation occurred to some degree in all dogs, with the most active dogs developing the largest seromas. A single latex drain was not adequate to control the fluid accumulation. The authors recommend placement of an additional latex drain in the inguinal area with appropriate bandaging to help obliterate dead space and minimize contamination. Alternatively, a closed suction device could be used to minimize the risk of ascending infection. Palpably normal muscle tone, visually normal abdominal wall contour, and histologic findings confirmed the clinical evidence of muscle viability. The persistence of function and integrity in the face of infection indicated that the lumbar external abdominal oblique myofascial island flap could be used in contaminated or infected wounds in which synthetic materials are considered contraindicated. The use of autogenous tissue for reconstruction provides an economical alternative to potentially costly synthetic prostheses.
Conclusion The lumbar external abdominal oblique myofascial island flap is easily harvested as a flap with a reliable neurovascular pedicle. The flap has a wide range of mobility and can survive in the face of infection. Additional inguinal drainage is necessary to control seroma formation.
References I . Voyles CR. Richardson JD. Bland KI, et a]. Emergency abdominal wall reconstruction with polypropylene mesh. Ann Surg I98 1: 194:2 19-223.
VASCULAR MYOFASCIAL FLAP
2. Luce EA, Hyde G, Gottlieb SE. Romm S. Total abdominal wall reconstruction. Arch Surg 1983; 1 1 8: 1446-1448. 3. Goris RJA. Ogilvie’s method applied to infected wound disruption. Arch Surg 1980; I15:1103-1107. 4. Sheridan R, D’Avis J , Seyfer AE, Quispe G. Massive abdominal wall desmoid tumor: Treatment by resection and abdominal wall reconstruction. Dis Colon and Rectum 1986; 29:5 18-52. 5 . Down NK, Falk RE. Makowa L. Excision ofabdominal wall tumors and reconstruction with marlex mesh. Can J Surg 1986: 29: 191193. 6. Weinstein LP, Kovachev D, Chaglassian T. Abdominal wall reconstruction. Scand J Plast Reconstr Surg 1986; 20: 109- I 13. 7. O’Hare PM. Leonard AG, Brennan MD. Experience with the tensor fascia latae free flap. Brit J Plast Surg 1983: 36:98-104. 8. Kaufman M, Weissberg D. Marlex mesh in giant ventral hernia repair. Israel J Med Sci 1980; 16:739-742. 9. Cady B, Brooke-Cowden GL. Repair of massive abdominal wall defects. Surg Clin N Am 1976:56:559-57. 10. Runnels CW, Trampel DW. Full-thickness thoracic and abdominal wall reconstruction in dogs using carbon/polycaprolactone composite. Vet Surg 1986: 15:363-368. I I . Usher FC. Ochsner JL. Marlex mesh: A new polyethylene mesh for replacing tissue defects. Surg Forum 1960; 10:319-323. 12. Brown GL, Richardson JD, Malangoni MA, et al. Comparison of prosthetic materials for abdominal wall reconstruction in the presence of contamination and infection. Ann Surg 1985: 201: 705-7 I I .
13. Hilbert BJ, Slatter DH, McDermott JD. Repair of a massive abdominal wall hernia in a horse using polypropylene mesh. Aust Vet J 1978: 54:588-59. 14. Nahai F. Silverton JS, Hill HL, Vasconez LO. The tensor fascia lata musculocutaneous flap. Ann Plast Surg 1978; 1:372-379. 15. Peled IJ, Kaplan HY, Herson M, Wexler MR. Tensor fascia lata musculocutaneous flap for abdominal wall reconstruction. Ann Plast Surg 1983; 11:141-143. 16. O’Hare PM, Leonard AG. Reconstruction of major abdominal wall defects using the tensor fascia lata myocutaneous flap. Brit J Plast Surg 1982:35:361-366. 7. Crowe DT, Archibald J. Abdominal wall and cavity. In: Archibald J. Calcott EJ, eds. Canine and Feline Surgery. 1st ed, vol. 1 Abdomen. Santa Barbara: American Veterinary Publications, Inc., 1987:79. 8. Pavletic MM, Kostolich M, Koblik P, Engler S. A comparison of the cutaneous trunci myocutaneous flap.and latissimus dorsi myocutaneous flap in the dog. Vet Surg 1987; 16:283-293. 19. Pavletic MM, Watters J, Henry RW, Nafe LA. Reverse saphenous conduit flap in the dog. J Am Vet Med Assn 1983; 182:380-389. 20. Weinstein MJ, Pavletic MM. Boudrieau RJ. Engler SJ. Caudal sartorius muscle flap in the dog. Vet Surg 1989; 18:286-291. 21. Gregory CR, Gourley IM, Koblik PD, Patz JD. Experimental definition of latissumus dorsi, gracilis, and rectus abdominus musculocutaneous flaps in the dog. Am J Vet Res 1988; 49:878-884. 22. Evans HE, Christiansen GC. eds. Miller’s Anatomy ofthe Dog. 2nd ed. Philadelphia: WB Saunders Co, 1979:728. 23. Lee AH, Swaim SF, Henderson RA. Surgical drainage. Comp Cont Ed 1986; 8:94-103.