- Email: [email protected]
Anatomical study of latissimus dorsi musculocutaneous flap vascular distribution
Journal of Plastic, Reconstructive & Aesthetic Surgery (2010) 63, 1091e1098
Anatomical study of latissimus dorsi musculocutaneous flap vascular distribution Koichi Watanabe a, Kensuke Kiyokawa a,*, Hideaki Rikimaru a, Noriyuki Koga a, Koh-ichi Yamaki b, Tsuyoshi Saga b a
Department of Plastic and Reconstructive Surgery and Maxillofacial Surgery, Kurume University School of Medicine, 67 Asahi-machi Kurume, Fukuoka 830-0011, Japan b Department of Anatomy, Kurume University School of Medicine, Fukuoka, Japan Received 17 July 2008; accepted 18 May 2009
KEYWORDS Latissimus dorsi musculocutaneous flap; Vascular territory; Angiosome; Choke vessels; Safety elevation area of flap
Summary Background: The objective of the current study is to elucidate the three-dimensional vascular distribution as far as the peripheral areas of a latissimus dorsi musculocutaneous flap and to establish a safe procedure for creating it. Methods: A lead oxide with gelatin-contrast agent was injected into fresh cadavers and the angiosomes in the muscle and skin were examined in detail. Results: In the muscle, three vascular territories were observed. The first vascular territory was formed by the thoracodorsal artery, the perforating branches of the ninth intercostal artery and those of the tenth intercostal artery located in the lateral part of the muscle. The second vascular territory was formed by the perforating branches of the tenth intercostal artery located in the medial part of the muscle, those of the 11th intercostal artery and the subcostal artery. The third vascular territory was formed by perforating branches of the first and second lumbar arteries. In the dorsal skin above the muscle, two vascular territories were observed. The first vascular territory was formed by perforating cutaneous branches of the thoracodorsal artery, perforating branches of the ninth through 11th intercostal arteries and the scapular circumflex artery. The second vascular territory was formed by perforating branches of the subcostal artery and the first and second lumbar arteries. Conclusions: When using a latissimus dorsi musculocutaneous flap with the thoracodorsal artery as a pedicle, the flap can be safely elevated as far as the inferior border of the 12th rib where perforating branches of the subcostal artery are distributed. At the same time, skin above the muscle can be safely harvested up to the iliac crest. It is essential, however, that the skin paddle includes perforating branches of the ninth intercostal artery or perforating branches of the 10th intercostal artery in the lateral part of the muscle. ª 2009 British Association of Plastic, Reconstructive and Aesthetic Surgeons. Published by Elsevier Ltd. All rights reserved.
* Corresponding author. Tel./fax: þ81 942 34 0834. E-mail address: [email protected] (K. Kiyokawa). 1748-6815/$ - see front matter ª 2009 British Association of Plastic, Reconstructive and Aesthetic Surgeons. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.bjps.2009.05.042
1092 Latissimus dorsi musculocutaneous flap is currently one of the most frequently used musculocutaneous flaps in reconstructive surgery. It has the advantages of enabling harvesting of large areas of muscle and skin with reliable blood supply, elevation of a scapular flap, anterior serratus muscle and scapular bone as a combined flap1,2 and elevation of an osteomusculocutaneous flap with attached rib bone.3,4 As a result, it is used for reconstruction of the head and neck,5,6 the chest7 and the upper arm,8 when used as a pedicled musculocutaneous flap, and various other areas of the body when used as a free musculocutaneous flap9 employing microvascular anastomosis. However, when harvesting a pedicled latissimus dorsi musculocutaneous flap extending as far as the peripheral region in the vicinity of the iliac crest, we experienced cases in which blood supply to the skin paddle and muscle at the periphery of the musculocutaneous flap was insufficient. Considering that musculocutaneous flaps have a very reliable blood supply, this is not the case at their peripheral parts. There are several reports regarding the blood supply of lattismus dorsi musculocutaneous flaps.10e14 However, there are no reports elucidating the three-dimensional route of blood flow from the thoracodorsal artery to the latissimus dorsi musculocutaneous flap or whether blood flow reaches the extremities of the muscle and skin paddle. Even now, when performing general microsurgery for reconstruction in cases of extensive loss on the limbs or trunk, in cases of failed microsurgery and cases of salvage operation, it is necessary to harvest latissimus dorsi musculocutaneous flaps up to the peripheral areas. In the current study, we conducted micro-angiography12,15 on fresh cadavers with an objective of elucidating the threedimensional vascular distribution of a latissimus dorsi musculocutaneous flap and establishing a reliable procedure for creating the flap.
Materials and methods The use of autopsy cadavers Prior to the sampling of tissue, informed consent was obtained from the attending physician and the bereaved family for use of the cadaver in the study. The procedures for tissue sampling from autopsy cases and imaging are as indicated below. 1) Full-thickness, one-piece samples of the latissimus dorsi muscle, the skin overlying it and thoraco-abdominal wall below the seventh rib were taken from the back of six sides of six fresh cadavers. 2) The harvested tissue from the back was maintained at a temperature of 37 C, about the same as body temperature, and physiological saline solution with heparin (1000 IU of heparin sodium mixed with 100 ml of physiological saline solution) at a temperature of 37 C was used to thoroughly flush the intravascular space. 3) Using a 50-ml syringe, a lead oxide with gelatincontrast agent with the same composition as that used by Rees et al.16 and at a temperature of 50 C was injected into the blood vessels with a moderate degree of pressure.
K. Watanabe et al.
Figure 1 Latissimus dorsi muscle angiogram findings. There are three vascular territories in the latissimus dorsi muscle, with choke vessels between each of them. (a) Horizontal branch of the thoracodorsal artery (b) Descending branch of the thoracodorsal artery (c) Perforating branches of the 9th intercostal artery in the medial part of the latissimus dorsi muscle (d) Perforating branch of the 9th intercostal artery in the lateral part of the latissimus dorsi muscle (e) Perforating branch of the 10th intercostal artery in the lateral part of the latissimus dorsi muscle (f) Perforating branches of the 10th intercostal artery (g) Perforating branches of the 11th intercostal artery (h) Perforating branches of a subcostal artery (i) Perforating branches of the 1st lumbar artery (j) Perforating branch of the 2nd lumbar artery.
4) After injecting the contrast agent, the tissue was stored for about 24 h at 4 C. When the contrast agent had solidified, radiographs of the latissimus dorsi muscle and the skin were taken separately. The radiographs were taken using computed radiography (Radiography system: Shimadzu X-ray High Voltage Genarator 150B-10, Image processing: Fuki film Computed Radiography FCR-HQ). Following examination of the sampled tissue, soft radiograph images of parts that required detailed examination were taken with a unit used for taking mammograms (Radiography system: Yokogawa Medical Systems Senograph 500t, Film: Kodak MIN-R 2000 (18 24 cm)).
Whole-body injection of the cadavers Whole-body injection of the contrast agent was performed on three fresh cadavers provided for anatomical training using the procedures indicated below. 1) With the aim of injecting contrast agent into the arterial system, incisions were made in the common carotid arteries and the femoral arteries on both sides of the cadavers, and 18Fr oral aspiration catheters (Terumo Safeed aspiration catheters) were placed in them. To reduce pressure on the venous system, 18Fr oral
Anatomical study of latissimus dorsi flap
1093
Figure 2 Dorsal skin angiogram findings. There are two angiosomes in the dorsal skin above the latissimus dorsi muscle, with choke vessels between them. (a) Circumflex scapular artery (b) Perforating branches of the thoracodorsal artery in the skin (c) Perforating branch of the 9th intercostal artery in the medial part of the latissimus dorsi muscle (d) Perforating branch of the 9th intercostal artery in the lateral part of the latissimus dorsi muscle (e) Perforating branch of the 10th intercostal artery in the lateral part of the latissimus dorsi muscle (f) Perforating branches of the 10th intercostal artery (g) Perforating branches of the 11th intercostal artery (h) Perforating branches of the subcostal artery (i) Perforating branches of the 1st lumbar artery (j) Perforating branch of the 2nd lumbar artery.
2)
3)
4) 5)
aspiration catheters were also placed in the femoral veins on both sides. Intra-arterial perfusion with 10 l of physiological saline solution with heparin (500 IU of heparin sodium mixed with 500 ml of physiological saline solution) at a temperature of 40 C was performed to thoroughly flush the intravascular space. Using a 50-ml syringe, a lead oxide with gelatincontrast agent with the same composition as that used by Rees et al.16 and at a temperature of 50 C was injected into the common carotid arteries and femoral arteries on both sides from four locations. The injected cadavers were kept in a 30% formalin bath for about 1 month to preserve the tissue. After completion of the above procedures, the same procedure used in the above-mentioned 1e4 was used to take radiographs of samples of the back of the chest wall and abdominal wall, including the latissimus dorsi muscle and the overlying dorsal skin.
Results Latissimus dorsi muscle blood circulation Three vascular territories were observed in the latissimus dorsi muscle (Figure 1). After entering the latissimus dorsi muscle from the axillary region, the thoracodorsal artery was seen to bifurcate.11 One of the bifurcations, a horizontal branch running parallel to the superior border of the latissimus dorsi muscle,
was directly connected as a result of perforating branches of the ninth intercostal artery in the medial part of the latissimus dorsi muscle and true anastomosis. The other bifurcation, a descending branch running parallel to the lateral border of the latissimus dorsi muscle, was directly connected as a result of perforating branches of the ninth and tenth intercostal arteries in the lateral part of the latissimus dorsi muscle and true anastomosis.14 In other words, the first vascular territory in the latissimus dorsi muscle was a vascular network formed by direct anastomosis of the thoracodorsal artery (horizontal branch, descending branch), perforating branches of the ninth intercostal artery and perforating branches of the tenth intercostal artery present in the lateral part of the latissimus dorsi muscle. The second vascular territory was a vascular network formed by direct anastomosis of perforating branches of the tenth intercostal artery present in the medial part of the latissimus dorsi muscle, perforating branches of the 11th intercostal artery and perforating branches of the subcostal artery.17e19 The third vascular territory was a sparse area of vascular network composed of perforating branches of the first and second lumbar arteries. Choke vessels were present between the abovementioned first and second, and second and third vascular territories.
Dorsal skin blood circulation Two vascular territories were observed in the dorsal skin overlying the latissimus dorsi muscle (Figure 2).
1094
K. Watanabe et al.
Figure 3 Three-dimensional vascular configuration of a latissimus dorsi musculocutaneous flap. TDA: Thoracodorsal artery Th9: Perforating branches of the 9th intercostal artery Th10: Perforating branches of the 10th intercostal artery Th11: Perforating branches of the 11th intercostal artery Th12: Perforating branches of a subcostal artery L1: Perforating branches of the 1st lumbar artery L2: Perforating branches of the 2nd lumbar artery L3: Perforating branches of the 3rd lumbar artery. (a) Lateral part of latissimus dorsi muscle and skin sagittal section angiogram findings. There are choke vessels between Th10, Th11 and Th12 in the latissimus dorsi muscle and between Th11 and Th12 in the skin. (b) Three-dimensional schema of an angiosome in a latissimus dorsi musculocutaneous flap.
The first vascular territory was a vascular network formed by direct anastomosis of perforating cutaneous branches of the thoracodorsal artery, perforating branches of the ninth, tenth and 11th intercostal arteries and the scapular circumflex artery.20 The second vascular territory was a vascular network formed by direct anastomosis of perforating branches of the subcostal artery and perforating branches of the first and second lumbar arteries. Choke vessels were present between the abovementioned first and second vascular territories. In the skin, caudally from the iliac crest where there is no latissimus dorsi muscle, there was another vascular territory completely distinct from the second vascular territory consisting of a vascular network formed by
anastomosis of the perforating branches of the third lumbar artery and perforating cutaneous branches from the gluteal muscle.
Latissimus dorsi musculocutaneous flap threedimensional vascular distribution Blood supply from the thoracodorsal artery to the first vascular territory in the latissimus dorsi muscle was primarily via perforating branches of the ninth intercostal artery and perforating branches of the tenth intercostal artery located in the lateral part of the latissimus dorsi muscle, with blood flowing directly to the first vascular territory in the skin. It is the first vascular territory of the
Anatomical study of latissimus dorsi flap
1095
Figure 4 Case 1 71-year-old male. Right pedicled latissimus dorsi musculocutaneous flap grafted to reconstruct the anterior wall of the cervical oesophagus after formation of an external pharyngeal fistula following surgery for hypopharyngeal cancer (a). Latissimus dorsi musculocutaneous flap created to include perforating branches of the 10th intercostals artery in the skin paddle, with the entire latissimus dorsi muscle harvested up to the vicinity of the iliac crest (b). Anterior wall of the cervical oesophagus reconstructed with a latissimus dorsi musculocutaneous flap skin paddle, with sufficient covering by the latissimus dorsi musculocutaneous flap muscle. Reticular, split-thickness skin graft also performed on the rear surface of the latissimus dorsi. After surgery, the skin paddle took completely but there was partial necrosis of the muscle and the split-thickness skin graft did not take (c, d).
latissimus dorsi muscle and the dorsal skin that form the 1st vascular territory of the latissimus dorsi musculocutaneous flap. The second vascular territory of a latissimus dorsi musculocutaneous flap was in the muscle from the edge of the first vascular territory to the inferior border of the 12th rib and in the skin extending to the iliac crest. Here, blood flow in the first vascular territory was through choke vessels located in the muscle and skin (Figure 3a and b).
Discussion The first report of a latissimus dorsi musculocutaneous flap was by Tansini21 in 1906. In 1912, d’Este22 reported the reconstruction of the chest wall using a latissimus dorsi musculocutaneous flap. With the consequent development of microsurgery, the same musculocutaneous flap was used as a free flap for reconstruction in various areas of the body.9 In recent years, ingenious new ways to use the flap have been developed, with methods of creating flaps for a wide range of applications, such as latissimus dorsi
perforator-based flaps reported by Angrigiani et al.23 and Spinelli et al.,24 and a split latissimus dorsi flap reported by Tobin et al.25 However, even with a latissimus dorsi musculocutaneous flap, which is often used since it is a safe flap with a stable blood supply, we have experienced a case where, despite complete adherence of the skin paddle, partial necrosis of just the underlying muscle occurred (Figure 4). From this experience, it was conjectured that even with a latissimus dorsi musculocutaneous flap, depending on how it is created, blood supply to the peripheral areas of the flap is not necessarily stable and that the regions of adhesion of muscle and skin paddle differ. In other words, even for a latissimus dorsi musculocutaneous flap, which is known to have an extremely stable blood supply, there is actually no clear understanding of the three-dimensional vascular distribution up to the peripheral area of the flap. In the current study, we examined in detail the three-dimensional vascular distribution of the thoracodorsal artery, which is a nutrient vessel for the latissimus dorsi muscle, and perforating branches of the costal, subcostal and lumbar
1096
K. Watanabe et al.
Figure 5 Case 2 72-year-old male. Reconstruction conducted with a pedicled latissimus dorsi musculocutaneous flap following the development of a cervical fistula after a total glossectomy, laryngectomy and mandibulectomy. The latissimus dorsi musculocutaneous flap skin paddle included perforating branches of the 10th intercostal artery present at the lateral border of the latissimus dorsi, and was long, extending to the lumbar region. Muscle was harvested up to the inferior border of the 12th rib (a, b). After surgery, both the musculocutaneous flap skin paddle and muscle took completely (c, d).
arteries in the latissimus dorsi muscle and dorsal skin, with the aim of establishing a method of creating safe latissimus dorsi musculocutaneous flaps. In a report related to the blood circulation of flaps by McGregor et al.26 in 1973, flaps were classified for the first time as being an axial pattern flap, which includes a nutrient artery, or a random pattern flap, which does not include one. The blood circulation of soft tissue was clarified in a report by Taylor et al.12 in 1987 that first introduced the concept of the angiosome (anatomical vascular territory), which receives nutrition through a single-source artery and is bordered by extremely fine blood vessels, called choke vessel, that gradually become narrower. Nakajima et al.27 as well reported on the changes in blood circulation, including choke vessels, that occur when elevating a flap. This referred to the elevation of an axial pattern flap, when the blood flow to the flap is from an axial nutrient vessel only, which is a pedicle, while the vascular pedicles of adjacent vascular territories are occluded. At this time, a pressure gradient is formed between the vascular territory supplied with nutrients by the axial nutrient vessel and adjacent vascular territories.
This pressure gradient causes the choke vessels between adjacent vascular territories to dilate, resulting in blood flowing from the axial nutrient vessel to the adjacent vascular territories, and the linking of blood flow. As a result, new blood circulation is created for the flap. This linking of blood flow occurs between the first vascular territory directly supplied by the axial nutrient vessel and, through choke vessels, the adjacent second vascular territory. However, linking of blood flow does not occur between the second and third vascular territories connected by choke vessels, and there is no blood flow from the axial nutrient vessel, which is a pedicle. In this case, the range of safe adhesion of the flap is up to the second vascular territory, with a high possibility of partial necrosis when elevating a flap that includes the third vascular territory.28,29 To clarify blood circulation when elevating a flap, the location of the choke vessels in the area used for the flap must be verified and the vascular territory must be clearly established. Furthermore, the location and blood circulation of choke vessels in the muscle, skin and subcutaneous tissue of the musculocutaneous flap and the connecting
Anatomical study of latissimus dorsi flap pathways between them must be verified three-dimensionally, and a ‘three-dimensional angiosome’ must be determined for the flap. This enables determination of the range for safe harvesting of the muscle and skin paddle for the musculocutaneous flap, and the perforating cutaneous branches (connecting pathways between skin and muscle) necessary in the skin paddle. Based on the above, Rikimaru et al.15 elucidated the three-dimensional vascular distribution of a pectoralis major musculocutaneous flap and a safe method for creating one. If the concept of a three-dimensional angiosome is applied to the latissimus dorsi musculocutaneous flap, then there are three vascular territories in a latissimus dorsi muscle, with the third vascular territory located caudally from the inferior border of the 12th rib. This is in accordance with the results we obtained with clinical cases in which necrosis of the muscle in the peripheral parts of latissimus dorsi musculocutaneous flaps occurred (Figure 4). Our results agreed with those reported by Taylor et al.12 These results point to the possibility that in actual clinical practice, latissimus dorsi muscle located caudally from the inferior border of the 12th rib does not have to be included in a musculocutaneous flap. Unlike the muscle, the skin above the latissimus dorsi has two vascular territories. Consequently, the skin paddle of a latissimus dorsi musculocutaneous flap can be safely harvested from the inferior edge of the second vascular territory, which extends up to the iliac crest. This is also in accordance with clinical cases in which there was safe adhesion of a long skin paddle harvested up to the iliac crest (Figure 5). The first vascular territory in the skin above the latissimus dorsi muscle is directly supplied with blood by the first vascular territory in the muscle, principally by perforating branches of the ninth intercostal artery, and perforating branches of the tenth intercostal artery located in the lateral part of the latissimus dorsi muscle. Together with the first vascular territory in the muscle, a first three-dimensional angiosome (vascular territory) of the latissimus dorsi musculocutaneous flap is created. Blood supply to this first vascular territory from the thoracodorsal artery flows through the choke vessels in the muscle and subcutaneous tissue into the second threedimensional angiosome (vascular territory) of the latissimus dorsi musculocutaneous flap (Figure 3b). Consequently, to ensure that blood flow reaches the second vascular territory of the latissimus dorsi musculocutaneous flap, particularly to the skin at the peripheral area over the iliac crest, the first essential step is to ensure that blood flow to the entire first vascular territory of the flap is sufficient. To achieve this, the skin paddle must include perforating branches of the ninth intercostal arteries and perforating branches of the tenth intercostal artery located in the lateral part of the latissimus dorsi muscle. It is important to secure a connecting pathway for blood flow from the first vascular territory of the muscle to that of the skin. If blood supply to this territory is sufficient, blood supply is possible through choke vessels to the extremities of the second vascular territory (Figure 5). From the above, important points concerning the actual creation of a safe latissimus dorsi musculocutaneous flap during clinical practice are as follows. When designing the skin paddle, use Doppler examination to verify that the
1097 perforating branches of the ninth intercostal artery and perforating branches of the tenth intercostal artery in the lateral part of the latissimus dorsi muscle are included, which is essential, and do not extend the range beyond the iliac crest. For the muscle, harvest up to the inferior border of the 12th rib (Figure 5). It is believed that adopting these procedures will assure the continued creation of safe latissimus dorsi musculocutaneous flaps with stable blood supply.
Conflict of interest None.
Funding None.
References 1. Harii K, Ono I, Ebihara S. Closure of total cheek defects with two combined myocutaneous free flaps. Arch Otolaryngol 1983;108:303e7. 2. Allen RJ, Dupin CL, Dreschnack PA, et al. The latissimus dorsi/scapular bone flap (the ‘‘latissimus/bone flap’’). Plast Reconstr Surg 1994;94:988e96. 3. Schmidt DR, Robson MC. One-stage composite reconstruction using the latissimus myocutaneous flap. Am J Surg 1982;144: 4470. 4. Maruyama Y, Urita Y, Ohnishi K. Rib-latissimus dorsi osteomyocutaneous flap in reconstruction of a mandibular defect. Br J Plast Surg 1985;38:234e7. 5. Quillen CG. Latissimus dorsi myocutaneous flaps in head and neck reconstruction. Plast Reconstr Surg 1979;63:664e70. 6. Chowdhuryn CR, McLean NR, Griffithes KH, et al. The repair of defects in the head and neck region with the latissimus dorsi myocutaneous flap. J Laryngol Otol 1988;102:1127e32. 7. Morgan RF, Edgerton MT, Wanebo HJ, et al. Reconstruction of full thickness chest wall defects. Ann Surg 1988;207:707e16. 8. Doi K, Sakai K, Ihara K, et al. Reinnervated free muscle transplantation for extreme reconstruction. Plast Reconstr Surg 1993;91:872e83. 9. Watson JS, Graig RDP, Orton CI. The free latissimus dorsi myocutaneous flap. Plast Reconstr Surg 1979;64:299e305. 10. Bartlett SP, May WM, Yaremchuk MJ. The latissimus dorsi muscle: a fresh cadaver study of the primary neurovascular pedicle. Plast Reconstr Surg 1981;67:631e6. 11. Roswell AR, Eisenberg N, Davies DM, et al. The anatomy of the thoracodorsal artery within the latissimus dorsi muscle. Br J Plast Surg 1986;39:206e9. 12. Taylor GI, Palmer JH. The vascular territories (angiosomes) of the body: experimental study and clinical applications. Br J Plast Surg 1987;40:113e41. 13. El-Maasarany SH, Sharaf E, Moustafa F, et al. Anatomical basis of latissimus dorsi myocutaneous flap: clinical applications. Surg Radiol Anat 1989;11:197e203. 14. Yang D, Morrris SF. Differences in intramuscular vascular connections of human and dog latissimus dorsi muscles. Ann Thorac Surg 1999;67:489e93. 15. Rikimaru H, Kiyokawa K, Inoue Y, et al. Three-dimensional anatomical vascular distribution in the pectoralis major myocutaneous flap. Plast Reconstr Surg 2005;115:1342e52. 16. Rees MJW, Taylor GI. A simplified lead oxide cadaver injection technique. Plast Reconstr Surg 1986;77:141e5.
1098 17. Beer GM, Lang A, Manestar M, et al. The bipedicle and bipartite latissimus dorsi free and perforator flap: an anatomic study. Plast Reconstr Surg 2008;118:1162e70. 18. Hamdi M, Spano A, van landuty K, et al. The lateral intercostal artery perforators: anatomical study and clinical application in breast surgery. Plast Reconstr Surg 2008;121:389e96. 19. Hamdi M, Van Landuty K, de Frene B, et al. The versatility of the inter-costal artery perforator (ICAP) flaps. J Plast Reconstr Aesthet Surg 2006;59:644e52. 20. Schaverien M, Saint-Cyr M, Arbique G, et al. Three- and fourdimensional arterial and venous anatomies of the thoracodorsal artery perforator flap. Plast Reconstr Surg 2008;121:1578e87. 21. Tansini I. Sopra il mino nuovo processo di amputazion della mammella. Gazz Med Ital 1906;57:141. 22. d’Este S. La technique de l’amputation la mammele pour carcinone mammaire. Rev Chir (Paris) 1912;45:164. 23. Angrigiani C, Grillio D, Siebert J. Latissimus dorsi musculocutaneous flap without muscle. Plast Reconstr Surg 1995;96: 1608e14.
K. Watanabe et al. 24. Spinelli HM, Fink JA, Muzaffar AR. The latissimus dorsi perforator-based faciocutaneous flap. Ann Plast Surg 1996;37: 500e6. 25. Tobin GR, Schusterman M, Peterson GH, et al. The intramuscular neurovascular anatomy of the latissimus dorsi muscle: the basis for splitting the flap. Plast Reconstr Surg 1981;67: 637e41. 26. McGregor IA, Morgan G. Axial and random pattern flaps. Br J Plast Surg 1973;26:202e13. 27. Nakajima H, Maruyama Y, Koda E. The definition of vascular territories with prostaglandin E1-the anterior chest, abdomen and thigh-inguinal region. Br J Plast Surg 1981;34: 258e63. 28. Callegari PR, Taylor GI, Caddy CM, et al. An anatomical review of the delay phenomenon: experimental studies. Plast Reconstr Surg 1992;89:397e407. 29. Taylor GI, Corlette RJ, Caddy CM, et al. An anatomical review of the delay phenomenon: clinical applications. Plast Reconstr Surg 1992;89:408e16.
Anatomical study of latissimus dorsi musculocutaneous flap vascular distribution Koichi Watanabe a, Kensuke Kiyokawa a,*, Hideaki Rikimaru a, Noriyuki Koga a, Koh-ichi Yamaki b, Tsuyoshi Saga b a
Department of Plastic and Reconstructive Surgery and Maxillofacial Surgery, Kurume University School of Medicine, 67 Asahi-machi Kurume, Fukuoka 830-0011, Japan b Department of Anatomy, Kurume University School of Medicine, Fukuoka, Japan Received 17 July 2008; accepted 18 May 2009
KEYWORDS Latissimus dorsi musculocutaneous flap; Vascular territory; Angiosome; Choke vessels; Safety elevation area of flap
Summary Background: The objective of the current study is to elucidate the three-dimensional vascular distribution as far as the peripheral areas of a latissimus dorsi musculocutaneous flap and to establish a safe procedure for creating it. Methods: A lead oxide with gelatin-contrast agent was injected into fresh cadavers and the angiosomes in the muscle and skin were examined in detail. Results: In the muscle, three vascular territories were observed. The first vascular territory was formed by the thoracodorsal artery, the perforating branches of the ninth intercostal artery and those of the tenth intercostal artery located in the lateral part of the muscle. The second vascular territory was formed by the perforating branches of the tenth intercostal artery located in the medial part of the muscle, those of the 11th intercostal artery and the subcostal artery. The third vascular territory was formed by perforating branches of the first and second lumbar arteries. In the dorsal skin above the muscle, two vascular territories were observed. The first vascular territory was formed by perforating cutaneous branches of the thoracodorsal artery, perforating branches of the ninth through 11th intercostal arteries and the scapular circumflex artery. The second vascular territory was formed by perforating branches of the subcostal artery and the first and second lumbar arteries. Conclusions: When using a latissimus dorsi musculocutaneous flap with the thoracodorsal artery as a pedicle, the flap can be safely elevated as far as the inferior border of the 12th rib where perforating branches of the subcostal artery are distributed. At the same time, skin above the muscle can be safely harvested up to the iliac crest. It is essential, however, that the skin paddle includes perforating branches of the ninth intercostal artery or perforating branches of the 10th intercostal artery in the lateral part of the muscle. ª 2009 British Association of Plastic, Reconstructive and Aesthetic Surgeons. Published by Elsevier Ltd. All rights reserved.
* Corresponding author. Tel./fax: þ81 942 34 0834. E-mail address: [email protected] (K. Kiyokawa). 1748-6815/$ - see front matter ª 2009 British Association of Plastic, Reconstructive and Aesthetic Surgeons. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.bjps.2009.05.042
1092 Latissimus dorsi musculocutaneous flap is currently one of the most frequently used musculocutaneous flaps in reconstructive surgery. It has the advantages of enabling harvesting of large areas of muscle and skin with reliable blood supply, elevation of a scapular flap, anterior serratus muscle and scapular bone as a combined flap1,2 and elevation of an osteomusculocutaneous flap with attached rib bone.3,4 As a result, it is used for reconstruction of the head and neck,5,6 the chest7 and the upper arm,8 when used as a pedicled musculocutaneous flap, and various other areas of the body when used as a free musculocutaneous flap9 employing microvascular anastomosis. However, when harvesting a pedicled latissimus dorsi musculocutaneous flap extending as far as the peripheral region in the vicinity of the iliac crest, we experienced cases in which blood supply to the skin paddle and muscle at the periphery of the musculocutaneous flap was insufficient. Considering that musculocutaneous flaps have a very reliable blood supply, this is not the case at their peripheral parts. There are several reports regarding the blood supply of lattismus dorsi musculocutaneous flaps.10e14 However, there are no reports elucidating the three-dimensional route of blood flow from the thoracodorsal artery to the latissimus dorsi musculocutaneous flap or whether blood flow reaches the extremities of the muscle and skin paddle. Even now, when performing general microsurgery for reconstruction in cases of extensive loss on the limbs or trunk, in cases of failed microsurgery and cases of salvage operation, it is necessary to harvest latissimus dorsi musculocutaneous flaps up to the peripheral areas. In the current study, we conducted micro-angiography12,15 on fresh cadavers with an objective of elucidating the threedimensional vascular distribution of a latissimus dorsi musculocutaneous flap and establishing a reliable procedure for creating the flap.
Materials and methods The use of autopsy cadavers Prior to the sampling of tissue, informed consent was obtained from the attending physician and the bereaved family for use of the cadaver in the study. The procedures for tissue sampling from autopsy cases and imaging are as indicated below. 1) Full-thickness, one-piece samples of the latissimus dorsi muscle, the skin overlying it and thoraco-abdominal wall below the seventh rib were taken from the back of six sides of six fresh cadavers. 2) The harvested tissue from the back was maintained at a temperature of 37 C, about the same as body temperature, and physiological saline solution with heparin (1000 IU of heparin sodium mixed with 100 ml of physiological saline solution) at a temperature of 37 C was used to thoroughly flush the intravascular space. 3) Using a 50-ml syringe, a lead oxide with gelatincontrast agent with the same composition as that used by Rees et al.16 and at a temperature of 50 C was injected into the blood vessels with a moderate degree of pressure.
K. Watanabe et al.
Figure 1 Latissimus dorsi muscle angiogram findings. There are three vascular territories in the latissimus dorsi muscle, with choke vessels between each of them. (a) Horizontal branch of the thoracodorsal artery (b) Descending branch of the thoracodorsal artery (c) Perforating branches of the 9th intercostal artery in the medial part of the latissimus dorsi muscle (d) Perforating branch of the 9th intercostal artery in the lateral part of the latissimus dorsi muscle (e) Perforating branch of the 10th intercostal artery in the lateral part of the latissimus dorsi muscle (f) Perforating branches of the 10th intercostal artery (g) Perforating branches of the 11th intercostal artery (h) Perforating branches of a subcostal artery (i) Perforating branches of the 1st lumbar artery (j) Perforating branch of the 2nd lumbar artery.
4) After injecting the contrast agent, the tissue was stored for about 24 h at 4 C. When the contrast agent had solidified, radiographs of the latissimus dorsi muscle and the skin were taken separately. The radiographs were taken using computed radiography (Radiography system: Shimadzu X-ray High Voltage Genarator 150B-10, Image processing: Fuki film Computed Radiography FCR-HQ). Following examination of the sampled tissue, soft radiograph images of parts that required detailed examination were taken with a unit used for taking mammograms (Radiography system: Yokogawa Medical Systems Senograph 500t, Film: Kodak MIN-R 2000 (18 24 cm)).
Whole-body injection of the cadavers Whole-body injection of the contrast agent was performed on three fresh cadavers provided for anatomical training using the procedures indicated below. 1) With the aim of injecting contrast agent into the arterial system, incisions were made in the common carotid arteries and the femoral arteries on both sides of the cadavers, and 18Fr oral aspiration catheters (Terumo Safeed aspiration catheters) were placed in them. To reduce pressure on the venous system, 18Fr oral
Anatomical study of latissimus dorsi flap
1093
Figure 2 Dorsal skin angiogram findings. There are two angiosomes in the dorsal skin above the latissimus dorsi muscle, with choke vessels between them. (a) Circumflex scapular artery (b) Perforating branches of the thoracodorsal artery in the skin (c) Perforating branch of the 9th intercostal artery in the medial part of the latissimus dorsi muscle (d) Perforating branch of the 9th intercostal artery in the lateral part of the latissimus dorsi muscle (e) Perforating branch of the 10th intercostal artery in the lateral part of the latissimus dorsi muscle (f) Perforating branches of the 10th intercostal artery (g) Perforating branches of the 11th intercostal artery (h) Perforating branches of the subcostal artery (i) Perforating branches of the 1st lumbar artery (j) Perforating branch of the 2nd lumbar artery.
2)
3)
4) 5)
aspiration catheters were also placed in the femoral veins on both sides. Intra-arterial perfusion with 10 l of physiological saline solution with heparin (500 IU of heparin sodium mixed with 500 ml of physiological saline solution) at a temperature of 40 C was performed to thoroughly flush the intravascular space. Using a 50-ml syringe, a lead oxide with gelatincontrast agent with the same composition as that used by Rees et al.16 and at a temperature of 50 C was injected into the common carotid arteries and femoral arteries on both sides from four locations. The injected cadavers were kept in a 30% formalin bath for about 1 month to preserve the tissue. After completion of the above procedures, the same procedure used in the above-mentioned 1e4 was used to take radiographs of samples of the back of the chest wall and abdominal wall, including the latissimus dorsi muscle and the overlying dorsal skin.
Results Latissimus dorsi muscle blood circulation Three vascular territories were observed in the latissimus dorsi muscle (Figure 1). After entering the latissimus dorsi muscle from the axillary region, the thoracodorsal artery was seen to bifurcate.11 One of the bifurcations, a horizontal branch running parallel to the superior border of the latissimus dorsi muscle,
was directly connected as a result of perforating branches of the ninth intercostal artery in the medial part of the latissimus dorsi muscle and true anastomosis. The other bifurcation, a descending branch running parallel to the lateral border of the latissimus dorsi muscle, was directly connected as a result of perforating branches of the ninth and tenth intercostal arteries in the lateral part of the latissimus dorsi muscle and true anastomosis.14 In other words, the first vascular territory in the latissimus dorsi muscle was a vascular network formed by direct anastomosis of the thoracodorsal artery (horizontal branch, descending branch), perforating branches of the ninth intercostal artery and perforating branches of the tenth intercostal artery present in the lateral part of the latissimus dorsi muscle. The second vascular territory was a vascular network formed by direct anastomosis of perforating branches of the tenth intercostal artery present in the medial part of the latissimus dorsi muscle, perforating branches of the 11th intercostal artery and perforating branches of the subcostal artery.17e19 The third vascular territory was a sparse area of vascular network composed of perforating branches of the first and second lumbar arteries. Choke vessels were present between the abovementioned first and second, and second and third vascular territories.
Dorsal skin blood circulation Two vascular territories were observed in the dorsal skin overlying the latissimus dorsi muscle (Figure 2).
1094
K. Watanabe et al.
Figure 3 Three-dimensional vascular configuration of a latissimus dorsi musculocutaneous flap. TDA: Thoracodorsal artery Th9: Perforating branches of the 9th intercostal artery Th10: Perforating branches of the 10th intercostal artery Th11: Perforating branches of the 11th intercostal artery Th12: Perforating branches of a subcostal artery L1: Perforating branches of the 1st lumbar artery L2: Perforating branches of the 2nd lumbar artery L3: Perforating branches of the 3rd lumbar artery. (a) Lateral part of latissimus dorsi muscle and skin sagittal section angiogram findings. There are choke vessels between Th10, Th11 and Th12 in the latissimus dorsi muscle and between Th11 and Th12 in the skin. (b) Three-dimensional schema of an angiosome in a latissimus dorsi musculocutaneous flap.
The first vascular territory was a vascular network formed by direct anastomosis of perforating cutaneous branches of the thoracodorsal artery, perforating branches of the ninth, tenth and 11th intercostal arteries and the scapular circumflex artery.20 The second vascular territory was a vascular network formed by direct anastomosis of perforating branches of the subcostal artery and perforating branches of the first and second lumbar arteries. Choke vessels were present between the abovementioned first and second vascular territories. In the skin, caudally from the iliac crest where there is no latissimus dorsi muscle, there was another vascular territory completely distinct from the second vascular territory consisting of a vascular network formed by
anastomosis of the perforating branches of the third lumbar artery and perforating cutaneous branches from the gluteal muscle.
Latissimus dorsi musculocutaneous flap threedimensional vascular distribution Blood supply from the thoracodorsal artery to the first vascular territory in the latissimus dorsi muscle was primarily via perforating branches of the ninth intercostal artery and perforating branches of the tenth intercostal artery located in the lateral part of the latissimus dorsi muscle, with blood flowing directly to the first vascular territory in the skin. It is the first vascular territory of the
Anatomical study of latissimus dorsi flap
1095
Figure 4 Case 1 71-year-old male. Right pedicled latissimus dorsi musculocutaneous flap grafted to reconstruct the anterior wall of the cervical oesophagus after formation of an external pharyngeal fistula following surgery for hypopharyngeal cancer (a). Latissimus dorsi musculocutaneous flap created to include perforating branches of the 10th intercostals artery in the skin paddle, with the entire latissimus dorsi muscle harvested up to the vicinity of the iliac crest (b). Anterior wall of the cervical oesophagus reconstructed with a latissimus dorsi musculocutaneous flap skin paddle, with sufficient covering by the latissimus dorsi musculocutaneous flap muscle. Reticular, split-thickness skin graft also performed on the rear surface of the latissimus dorsi. After surgery, the skin paddle took completely but there was partial necrosis of the muscle and the split-thickness skin graft did not take (c, d).
latissimus dorsi muscle and the dorsal skin that form the 1st vascular territory of the latissimus dorsi musculocutaneous flap. The second vascular territory of a latissimus dorsi musculocutaneous flap was in the muscle from the edge of the first vascular territory to the inferior border of the 12th rib and in the skin extending to the iliac crest. Here, blood flow in the first vascular territory was through choke vessels located in the muscle and skin (Figure 3a and b).
Discussion The first report of a latissimus dorsi musculocutaneous flap was by Tansini21 in 1906. In 1912, d’Este22 reported the reconstruction of the chest wall using a latissimus dorsi musculocutaneous flap. With the consequent development of microsurgery, the same musculocutaneous flap was used as a free flap for reconstruction in various areas of the body.9 In recent years, ingenious new ways to use the flap have been developed, with methods of creating flaps for a wide range of applications, such as latissimus dorsi
perforator-based flaps reported by Angrigiani et al.23 and Spinelli et al.,24 and a split latissimus dorsi flap reported by Tobin et al.25 However, even with a latissimus dorsi musculocutaneous flap, which is often used since it is a safe flap with a stable blood supply, we have experienced a case where, despite complete adherence of the skin paddle, partial necrosis of just the underlying muscle occurred (Figure 4). From this experience, it was conjectured that even with a latissimus dorsi musculocutaneous flap, depending on how it is created, blood supply to the peripheral areas of the flap is not necessarily stable and that the regions of adhesion of muscle and skin paddle differ. In other words, even for a latissimus dorsi musculocutaneous flap, which is known to have an extremely stable blood supply, there is actually no clear understanding of the three-dimensional vascular distribution up to the peripheral area of the flap. In the current study, we examined in detail the three-dimensional vascular distribution of the thoracodorsal artery, which is a nutrient vessel for the latissimus dorsi muscle, and perforating branches of the costal, subcostal and lumbar
1096
K. Watanabe et al.
Figure 5 Case 2 72-year-old male. Reconstruction conducted with a pedicled latissimus dorsi musculocutaneous flap following the development of a cervical fistula after a total glossectomy, laryngectomy and mandibulectomy. The latissimus dorsi musculocutaneous flap skin paddle included perforating branches of the 10th intercostal artery present at the lateral border of the latissimus dorsi, and was long, extending to the lumbar region. Muscle was harvested up to the inferior border of the 12th rib (a, b). After surgery, both the musculocutaneous flap skin paddle and muscle took completely (c, d).
arteries in the latissimus dorsi muscle and dorsal skin, with the aim of establishing a method of creating safe latissimus dorsi musculocutaneous flaps. In a report related to the blood circulation of flaps by McGregor et al.26 in 1973, flaps were classified for the first time as being an axial pattern flap, which includes a nutrient artery, or a random pattern flap, which does not include one. The blood circulation of soft tissue was clarified in a report by Taylor et al.12 in 1987 that first introduced the concept of the angiosome (anatomical vascular territory), which receives nutrition through a single-source artery and is bordered by extremely fine blood vessels, called choke vessel, that gradually become narrower. Nakajima et al.27 as well reported on the changes in blood circulation, including choke vessels, that occur when elevating a flap. This referred to the elevation of an axial pattern flap, when the blood flow to the flap is from an axial nutrient vessel only, which is a pedicle, while the vascular pedicles of adjacent vascular territories are occluded. At this time, a pressure gradient is formed between the vascular territory supplied with nutrients by the axial nutrient vessel and adjacent vascular territories.
This pressure gradient causes the choke vessels between adjacent vascular territories to dilate, resulting in blood flowing from the axial nutrient vessel to the adjacent vascular territories, and the linking of blood flow. As a result, new blood circulation is created for the flap. This linking of blood flow occurs between the first vascular territory directly supplied by the axial nutrient vessel and, through choke vessels, the adjacent second vascular territory. However, linking of blood flow does not occur between the second and third vascular territories connected by choke vessels, and there is no blood flow from the axial nutrient vessel, which is a pedicle. In this case, the range of safe adhesion of the flap is up to the second vascular territory, with a high possibility of partial necrosis when elevating a flap that includes the third vascular territory.28,29 To clarify blood circulation when elevating a flap, the location of the choke vessels in the area used for the flap must be verified and the vascular territory must be clearly established. Furthermore, the location and blood circulation of choke vessels in the muscle, skin and subcutaneous tissue of the musculocutaneous flap and the connecting
Anatomical study of latissimus dorsi flap pathways between them must be verified three-dimensionally, and a ‘three-dimensional angiosome’ must be determined for the flap. This enables determination of the range for safe harvesting of the muscle and skin paddle for the musculocutaneous flap, and the perforating cutaneous branches (connecting pathways between skin and muscle) necessary in the skin paddle. Based on the above, Rikimaru et al.15 elucidated the three-dimensional vascular distribution of a pectoralis major musculocutaneous flap and a safe method for creating one. If the concept of a three-dimensional angiosome is applied to the latissimus dorsi musculocutaneous flap, then there are three vascular territories in a latissimus dorsi muscle, with the third vascular territory located caudally from the inferior border of the 12th rib. This is in accordance with the results we obtained with clinical cases in which necrosis of the muscle in the peripheral parts of latissimus dorsi musculocutaneous flaps occurred (Figure 4). Our results agreed with those reported by Taylor et al.12 These results point to the possibility that in actual clinical practice, latissimus dorsi muscle located caudally from the inferior border of the 12th rib does not have to be included in a musculocutaneous flap. Unlike the muscle, the skin above the latissimus dorsi has two vascular territories. Consequently, the skin paddle of a latissimus dorsi musculocutaneous flap can be safely harvested from the inferior edge of the second vascular territory, which extends up to the iliac crest. This is also in accordance with clinical cases in which there was safe adhesion of a long skin paddle harvested up to the iliac crest (Figure 5). The first vascular territory in the skin above the latissimus dorsi muscle is directly supplied with blood by the first vascular territory in the muscle, principally by perforating branches of the ninth intercostal artery, and perforating branches of the tenth intercostal artery located in the lateral part of the latissimus dorsi muscle. Together with the first vascular territory in the muscle, a first three-dimensional angiosome (vascular territory) of the latissimus dorsi musculocutaneous flap is created. Blood supply to this first vascular territory from the thoracodorsal artery flows through the choke vessels in the muscle and subcutaneous tissue into the second threedimensional angiosome (vascular territory) of the latissimus dorsi musculocutaneous flap (Figure 3b). Consequently, to ensure that blood flow reaches the second vascular territory of the latissimus dorsi musculocutaneous flap, particularly to the skin at the peripheral area over the iliac crest, the first essential step is to ensure that blood flow to the entire first vascular territory of the flap is sufficient. To achieve this, the skin paddle must include perforating branches of the ninth intercostal arteries and perforating branches of the tenth intercostal artery located in the lateral part of the latissimus dorsi muscle. It is important to secure a connecting pathway for blood flow from the first vascular territory of the muscle to that of the skin. If blood supply to this territory is sufficient, blood supply is possible through choke vessels to the extremities of the second vascular territory (Figure 5). From the above, important points concerning the actual creation of a safe latissimus dorsi musculocutaneous flap during clinical practice are as follows. When designing the skin paddle, use Doppler examination to verify that the
1097 perforating branches of the ninth intercostal artery and perforating branches of the tenth intercostal artery in the lateral part of the latissimus dorsi muscle are included, which is essential, and do not extend the range beyond the iliac crest. For the muscle, harvest up to the inferior border of the 12th rib (Figure 5). It is believed that adopting these procedures will assure the continued creation of safe latissimus dorsi musculocutaneous flaps with stable blood supply.
Conflict of interest None.
Funding None.
References 1. Harii K, Ono I, Ebihara S. Closure of total cheek defects with two combined myocutaneous free flaps. Arch Otolaryngol 1983;108:303e7. 2. Allen RJ, Dupin CL, Dreschnack PA, et al. The latissimus dorsi/scapular bone flap (the ‘‘latissimus/bone flap’’). Plast Reconstr Surg 1994;94:988e96. 3. Schmidt DR, Robson MC. One-stage composite reconstruction using the latissimus myocutaneous flap. Am J Surg 1982;144: 4470. 4. Maruyama Y, Urita Y, Ohnishi K. Rib-latissimus dorsi osteomyocutaneous flap in reconstruction of a mandibular defect. Br J Plast Surg 1985;38:234e7. 5. Quillen CG. Latissimus dorsi myocutaneous flaps in head and neck reconstruction. Plast Reconstr Surg 1979;63:664e70. 6. Chowdhuryn CR, McLean NR, Griffithes KH, et al. The repair of defects in the head and neck region with the latissimus dorsi myocutaneous flap. J Laryngol Otol 1988;102:1127e32. 7. Morgan RF, Edgerton MT, Wanebo HJ, et al. Reconstruction of full thickness chest wall defects. Ann Surg 1988;207:707e16. 8. Doi K, Sakai K, Ihara K, et al. Reinnervated free muscle transplantation for extreme reconstruction. Plast Reconstr Surg 1993;91:872e83. 9. Watson JS, Graig RDP, Orton CI. The free latissimus dorsi myocutaneous flap. Plast Reconstr Surg 1979;64:299e305. 10. Bartlett SP, May WM, Yaremchuk MJ. The latissimus dorsi muscle: a fresh cadaver study of the primary neurovascular pedicle. Plast Reconstr Surg 1981;67:631e6. 11. Roswell AR, Eisenberg N, Davies DM, et al. The anatomy of the thoracodorsal artery within the latissimus dorsi muscle. Br J Plast Surg 1986;39:206e9. 12. Taylor GI, Palmer JH. The vascular territories (angiosomes) of the body: experimental study and clinical applications. Br J Plast Surg 1987;40:113e41. 13. El-Maasarany SH, Sharaf E, Moustafa F, et al. Anatomical basis of latissimus dorsi myocutaneous flap: clinical applications. Surg Radiol Anat 1989;11:197e203. 14. Yang D, Morrris SF. Differences in intramuscular vascular connections of human and dog latissimus dorsi muscles. Ann Thorac Surg 1999;67:489e93. 15. Rikimaru H, Kiyokawa K, Inoue Y, et al. Three-dimensional anatomical vascular distribution in the pectoralis major myocutaneous flap. Plast Reconstr Surg 2005;115:1342e52. 16. Rees MJW, Taylor GI. A simplified lead oxide cadaver injection technique. Plast Reconstr Surg 1986;77:141e5.
1098 17. Beer GM, Lang A, Manestar M, et al. The bipedicle and bipartite latissimus dorsi free and perforator flap: an anatomic study. Plast Reconstr Surg 2008;118:1162e70. 18. Hamdi M, Spano A, van landuty K, et al. The lateral intercostal artery perforators: anatomical study and clinical application in breast surgery. Plast Reconstr Surg 2008;121:389e96. 19. Hamdi M, Van Landuty K, de Frene B, et al. The versatility of the inter-costal artery perforator (ICAP) flaps. J Plast Reconstr Aesthet Surg 2006;59:644e52. 20. Schaverien M, Saint-Cyr M, Arbique G, et al. Three- and fourdimensional arterial and venous anatomies of the thoracodorsal artery perforator flap. Plast Reconstr Surg 2008;121:1578e87. 21. Tansini I. Sopra il mino nuovo processo di amputazion della mammella. Gazz Med Ital 1906;57:141. 22. d’Este S. La technique de l’amputation la mammele pour carcinone mammaire. Rev Chir (Paris) 1912;45:164. 23. Angrigiani C, Grillio D, Siebert J. Latissimus dorsi musculocutaneous flap without muscle. Plast Reconstr Surg 1995;96: 1608e14.
K. Watanabe et al. 24. Spinelli HM, Fink JA, Muzaffar AR. The latissimus dorsi perforator-based faciocutaneous flap. Ann Plast Surg 1996;37: 500e6. 25. Tobin GR, Schusterman M, Peterson GH, et al. The intramuscular neurovascular anatomy of the latissimus dorsi muscle: the basis for splitting the flap. Plast Reconstr Surg 1981;67: 637e41. 26. McGregor IA, Morgan G. Axial and random pattern flaps. Br J Plast Surg 1973;26:202e13. 27. Nakajima H, Maruyama Y, Koda E. The definition of vascular territories with prostaglandin E1-the anterior chest, abdomen and thigh-inguinal region. Br J Plast Surg 1981;34: 258e63. 28. Callegari PR, Taylor GI, Caddy CM, et al. An anatomical review of the delay phenomenon: experimental studies. Plast Reconstr Surg 1992;89:397e407. 29. Taylor GI, Corlette RJ, Caddy CM, et al. An anatomical review of the delay phenomenon: clinical applications. Plast Reconstr Surg 1992;89:408e16.