- Email: [email protected]
Applied anatomy of the latissimus dorsi free flap for refinement in one-stage facial reanimation
Journal of Plastic, Reconstructive & Aesthetic Surgery (2011) 64, 1417e1423
Applied anatomy of the latissimus dorsi free flap for refinement in one-stage facial reanimation* L.D. Ferguson a,*, T. Paterson a, F. Ramsay a, K. Arrol a, J. Dabernig b, J. Shaw-Dunn a, S. Morley b a
Department of Human Anatomy, Thomson Building, University of Glasgow, Glasgow G12 8QQ, UK Canniesburn Plastic Surgery Unit, Glasgow Royal Infirmary, Jubilee Building, 84 Castle Street, Glasgow G4 0SF, UK
b
Received 4 December 2010; accepted 7 June 2011
KEYWORDS Latissimus dorsi flap; Segmental; Facial reanimation; Thoracodorsal pedicle
Summary Background: The face can be reanimated after long-term paralysis by free microneurovascular tissue transfer. Flaps from gracilis and pectoralis minor usually require a twostage procedure with a cross-face nerve graft. Latissimus dorsi has a much longer muscular nerve, the thoracodorsal nerve, which could avoid the need for a second cross-face nerve graft. Our hypothesis is that the neurovascular pedicles of small segments of latissimus dorsi would be long enough to reach the opposite side of the face and to provide a reliable blood and nerve supply to the flaps. Method: To test this hypothesis the thoracodorsal pedicle and its primary branches were dissected in eleven embalmed cadavers. The segmental vessels and nerves were then traced in a series of simulated flaps approximately 8e10 cm 2e3 cm by micro-dissection, tissue clearing and histology. Results: The thoracodorsal pedicle is 10e14 cm long to where it enters the muscle, and with intra-muscular dissection small chimeric muscle segments 8e10 cm 2e3 cm can be raised with a clearly defined neurovascular supply. Using micro-dissection the neurovascular pedicle can be lengthened to reach across the face. Segmental arteries and nerves extended to the distal end of all the flaps examined. Artery, vein and nerve run together and are of substantial diameter. Conclusion: Small muscle segments of latissimus dorsi can be raised on long neurovascular pedicles. The vessels and nerves are substantial and the likelihood of surgical complications such as flap necrosis and functional disuse on transplantation appear low. Although in our
* Presentations: Research related to this article has been presented by the corresponding author LD Ferguson at the Anatomical Society of Great Britain and Ireland (ASGBI) Winter Meeting January 2009 and by T Paterson at the British Association of Clinical Anatomists (BACA) Winter Meeting December 2008, at which it was awarded the Conrad Lewin prize. * Corresponding author. E-mail address: [email protected] (L.D. Ferguson).
1748-6815/$ - see front matter ª 2011 British Association of Plastic, Reconstructive and Aesthetic Surgeons. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.bjps.2011.06.013
1418
L.D. Ferguson et al. opinion the use of cross-face nerve grafts and transfer of smaller muscle flaps remains the gold standard in facial reanimation in straightforward cases, the micro-dissected latissimus dorsi flap is a useful option in complex cases of facial reanimation. Clinical Application: Facial reanimation using micro-dissected segments of latissimus dorsi has been performed in four complex cases of facial paralysis. ª 2011 British Association of Plastic, Reconstructive and Aesthetic Surgeons. Published by Elsevier Ltd. All rights reserved.
Introduction Facial paralysis is disfiguring, with emotional expression, chewing, speech, and eye closure affected and asymmetry of the face at rest. Surgical reanimation techniques for irreversible facial paralysis mostly focus on attempts to restore voluntary facial movement and emotional expression i.e. to recreate the smile. Irreversible can be taken to mean long standing facial paralysis which is unlikely to recover either due to the nature of the injury or the time elapsed. Several potential donor muscles have been used, the commonest muscles being the gracilis and pectoralis minor flaps, usually implanted as a second stage following a cross-face nerve graft.1,2 In complex cases or where a single stage procedure is desired a larger muscle with a longer pedicle may be used and the most frequently used in this circumstance is latissimus dorsi. This is due to its long neurovascular pedicle and low donor site morbidity.3 The latissimus dorsi is particularly useful where a long pedicle is required due to vessel depletion in the neck or where a cross-face nerve graft cannot be used because a one-stage reconstruction is required. Latissimus dorsi is a type V muscle in the classification of Mathes and Nahai (1981) with one dominant vascular pedicle from the thoracodorsal artery at its insertion, and several secondary perforator pedicles arising from the 9th, 10th, and 11th terminal intercostal and 1st, 2nd, and 3rd lumbar arteries, at its origin.4 Beer (2006) has argued the merits of a “bipedicled” flap based on this dual supply5 but the majority of latissimus dorsi free flaps are still designed solely on the dominant thoracodorsal pedicle.6 In its entirety the muscle is far too large to be transplanted to the face as a whole and so a basis has been sought for splitting the muscle into viable smaller segments with a long pedicle. Initial studies by Bartlett (1981) and Tobin (1981) revealed primary bifurcation of the thoracodorsal vessels into medial and lateral branches.7,8 The key finding however was the subsequent discovery of 4e6 segmental (secondary) arteries and nerves by Zhao et al. (1993) and Wong et al. (2007).9,10 These findings suggested that latissimus dorsi could be split into 4e6 smaller segmental muscle flaps as independent, functional units for transplantation to the face.11 A recent paper has looked at the latissimus dorsi pedicle in relation to its three dimensional blood supply both to the muscle and the overlying skin, showing a consistent branching pattern in three separate vascular territories with choke vessels joining these areas.12 For segmental muscle flaps to be successful as a one-stage facial reanimation technique the neurovascular pedicle of latissimus dorsi must be long enough to reach across the face, and the segmental flaps must have an adequate blood and nerve supply throughout their length. Several authors have
used the latissimus dorsi muscle in this way with success. We wished to determine more precisely the anatomic basis of the muscle as the pedicle splits and passes from proximal to distal. In particular we wished to determine for how far along its length is the segmental neurovascular supply of the latissimus dorsi muscle predictable. To this end several gross dissections of muscle were made; histology was used to determine the microscopic arrangement of vessels and nerve and simulated reanimation operations with segmental flaps of latissimus dorsi were undertaken. The knowledge gained by these studies has been applied in four cases of complex facial reanimation. These cases demonstrated the need either for a long vascular pedicle required because of previous surgery or a long nerve to reach the contralateral side of the face.
Materials and methods Tissue used in this study was obtained from eleven cadavers donated to the Department of Anatomy, University of Glasgow. These had been prepared by common carotid artery injection of embalming fluid followed after 48 h by a threepart dilution of ammonia and, finally, a mixture of India ink and latex rubber. Human tissue is held under the provisions of the Anatomy Act (1984) as revised in the Human Tissue (Scotland) Act 2006, which includes ethical approval.
Gross- and micro-dissection A mid-axillary incision was used to locate the distal third of the axillary artery from which the subscapular artery arose. The constituents of the neurovascular pedicle, the thoracodorsal artery, vein and nerve were identified and traced to their entry into latissimus dorsi at the neurovascular hilum. In each of eleven latissimus dorsi muscles, the senior author fashioned four to five segments approximately 8e10 cm 2e3 cm, based on the visible branches of the thoracodorsal artery and nerve, as at operation. These pieces of muscle were dissected distal to the point of insertion of the neurovascular pedicle meaning muscle flap segments were raised from approximately 12e14 cm to 20e24 cm from the origin of the muscle in the axilla (Figures 1 and 2). The flaps were dissected with the aid of a dissecting microscope to trace the segmental artery along the full length of each flap. The main focus of the study was on raising lateral segmental flaps of muscle placed distally, as would be used in reanimation surgery.
Tissue clearing To trace the smaller branches of the arteries the muscle was cleared. In this technique, the muscle is fully
Applied anatomy of the latissimus dorsi free flap
1419
Simulated facial reanimation
Figure 1 Deep surface of latissimus dorsi muscle surgically split into 4 segmental flaps based on its segmental branches (white asterisks). Segments 1 and 2 are the most lateral and readily exposed at operation. The thoracodorsal pedicle, along with its lateral and medial branches, is also visible.
dehydrated in alcohol then immersed in methyl salicylate which makes the tissue translucent.13 In the final result, the arteries filled with Indian ink appear jet black on a golden yellow background in which the main tissue elements are still recognisable.
Histology Full thickness transverse sections were taken at 4 cm intervals along the segmental muscle flaps and processed for routine histology. Sections were taken at specified points on the pathway of the neurovascular pedicle from proximal to distal with particular emphasis on the pedicle arrangement distally. These flaps would mimic the piece of muscle which would be used in facial reanimation using a pedicle sufficiently long to reach the contralateral cheek. Arterial and nerve diameters in the sections were measured with a Zeiss AxioSkop2 Mot Plus microscope with Axiocam for digital capture and Zeiss AxioVision software version 4.7, (Carl Zeiss MicroImaging GmbH, Gottingen, Germany).
At the simulated operation, two segmental flaps of latissimus dorsi measuring 8 cm in length and 3 cm in width were raised on extended pedicles which had been lengthened with intra-muscular dissection. They were then attached into position medially at the modiolus and angled laterally onto the temporal fascia. The thoracodorsal nerve was passed through the buccal cavity of the cheek with a pair of forceps into the mouth. From the mouth, the nerve was tunnelled through the upper lip and fed into the contralateral side of the face. Here the thoracodorsal nerve met the pre-dissected branches of the facial nerve. This mock procedure confirmed that the nerve, once appropriately dissected, is long enough to pass across the face for coaptation to the contralateral facial nerve, exposed through a pre-auricular incision.
Clinical examples Since the anatomical side of the study was completed, a segmental latissimus dorsi flap has been used in complex facial reanimation in four cases. In each case the anatomical knowledge gained from the cadaver studies was used to expedite micro-dissection of precisely formed segments of the latissimus dorsi muscle, on a long pedicle. Flaps were raised using a lateral incision in the axilla with the patient in the supine position.
Case one A 44 year old female presented to the facial palsy clinic with a dense left facial paralysis secondary to radical excisional surgery followed by chemoradiotherapy to treat a recurrent parotid adenocarcinoma. She already had a free flap (myocutaneous rectus flap) on her left cheek and the temporalis muscle had been excised along with the facial nerve. It was decided that a one-stage free tissue transfer using a latissimus dorsi muscle would give the best chance of both an improved static position and restoration of smile. Due to her previous surgery this would have to be motored onto the contralateral facial nerve and recipient vessels would be located deep in the neck, meaning a long vascular pedicle was required. Two chimeric muscle paddles of 8 cm 3 cm were raised with a 15 cm length of nerve, and inset into the left cheek in front of the pre-existing free flap. The thoracodorsal nerve was coapted onto a suitable branch of the facial nerve on the contralateral side, exposed through a pre-auricular incision. At 18 months follow up this has given an improved static position where the position of the modiolus has been lifted 9 mm to leave it symmetrical with the normal side. Dynamic correction of the palsy which is spontaneous with the patient’s smile, has also been achieved with 5 mm of modiolus excursion, although the smile is not fully symmetrical.
Case two
Figure 2 Latissimus dorsi has been surgically split into 5 segmental flaps in this case based upon the 5 segmental vessels and nerves (A to E).
A 51 year old female presented to the facial palsy clinic with a dense left facial paralysis secondary to radical excisional surgery and radiotherapy to treat a maxillary carcinoma. She had been left with a mandibular defect due
1420
L.D. Ferguson et al. of 10 cm 3 cm was raised on a micro-dissected intramuscular pedicle 10 cm distal to the main bulk of the flap. The nerve was crossed to the contralateral facial nerve and coapted to a suitable buccal branch. At 6 months follow up the procedure has resulted in an improved static position with 5 mm of modiolus elevation and 3 mm of modiolus excursion when smiling.
Case three
Figure 3 Deep surface of latissimus dorsi still attached to the body. The lateral and medial branches of the thoracodorsal pedicle can be seen at the neurovascular hilum. From these arise the segmental vessels and nerves upon which the muscle flaps are based (Figure 1). Secondary perforators can also be seen at its origin.
to severe osteo-radionecrosis. The original excisional surgeon required to place a muscle free flap to correct the mandibular soft tissue defect. It was decided therefore to raise a chimeric latissimus dorsi, with the main bulk of the flap being used to correct this defect but with a segmental muscle paddle being used for facial reanimation. A segment
A 32 year old female presented to the facial palsy clinic having had previous facial reanimation surgery at another unit with the insertion of a free rectus flap performed in another unit. Her initial palsy was related to excision of a vestibular schwannoma. The muscle insertion had dehisced both proximally and distally meaning that the muscular contraction of the muscle was producing a bunched up appearance. One attempt to reposition the existing muscle was unsuccessful and following this surgery there was only a flicker of muscle movement. It was decided to remove the existing muscle and replace it with a second muscle transfer. Due to her previous surgery we knew a long vascular pedicle would be required so a segmental latissimus dorsi flap was raised. As the crossface nerve graft was not seen to be working strongly the thoracodorsal nerve was tunnelled to the contralateral side to provide a fresh neural input. The full length of the pedicle was required as the lingual artery was used as a recipient vessel. At 13 months follow up the operation has successfully improved her static position by a 6 mm lift at the modiolus and she now has spontaneous movement of the modiolus of 5 mm when smiling.
Case four A 44 year old male with bilateral facial paralysis secondary to acoustic neuroma surgery presented to clinic complaining of
Figure 4 Another latissimus dorsi flap still attached to the body by the thoracodorsal (TD) pedicle. In this case, the pedicle divides into 5 segmental branches upon which the flaps in Figure 2 are based. A is the lateral branch.
Figure 5 Latissimus dorsi after clearing. The arteries injected with Indian Ink appear black. There are three substantial arteries which spread out parallel to the radiating muscle fasciculi. The lateral branch gives two subdivisions (A & B) giving in all four segmental arteries AeD. Anastomoses can be traced with secondary perforators to the right of the figure.
Applied anatomy of the latissimus dorsi free flap Table 1
1421
Segmental artery and nerve diameters at proximal, middle, and distal aspects of a typical segmental flap.
Distance along muscle segment:
Diameter of segmental artery (mm):
Diameter of segmental nerve (mm):
Proximal (0 cm) Middle (4 cm) Distal (8 cm)
1.74 0.80 0.62 and 0.41
0.53 and 0.49 0.34 and 0.28 0.45
asymmetry and complete lack of facial movement on the right side. Many years previously he had had an attempt at facial reanimation using a free gracilis muscle transfer which had not been successful in re-animating his face. He had bilateral fascia lata static supports added at an earlier procedure. The patient received a segmentally split latissimus dorsi flap motored onto the nerve to masseter as a one-stage procedure. The pedicle was dissected into the muscle to allow a length of pedicle to be harvested sufficient to reach deep into the neck to obtain suitable vessels for micro-anastamosis which was required as the facial vessels in the cheek had been utilised previously for microsurgery. Although the superficial temporal vessels could potentially have been used as recipient vessels due to the patient’s previous procedures it was not felt that these would have been reliable. At 6 months follow up the combination of fascia lata supports and free tissue transfer have successfully improved his static position by a 20 mm lift at the modiolus and he now has movement of the modiolus of 7 mm when smiling.
Results The latissimus dorsi is a large triangular superficial muscle of the back with a narrow insertion into the intertubercular groove of the humerus. The lower fibres are near vertical, the middle oblique, and the upper fibres almost horizontal. The lateral border is readily exposed at dissection and at operation. The thoracodorsal artery divides as it approaches the neurovascular hilum on the deep surface of
Figure 6 Miller and Van Gieson stained section taken from proximal part of a typical segmental flap. The segmental artery (A) with two accompanying segmental veins (V) and nerves (N) are seen surrounded by connective tissue in a neurovascular bundle.
the muscle. The usual pattern is two divisions, lateral and medial (Figure 3), but occasionally there are three. Within the muscle, segmental branches spread out parallel to the radiating muscle fasciculi. These branches vary in number and diameter but in our series of eleven cadaver dissections there were never fewer than three substantial branches and often four or five (Figures 1, 4, and 5). The lateral branch follows the lateral border of the muscle at a distance of 1e2 cm and often divides into two major subdivisions (Figure 5). There are large anastomoses with the secondary perforators (Figure 5). The thoracodorsal nerve divides proximal to the artery. The branches usually follow the arteries in their radiating course, but occasionally the main trunks may diverge. The total length of the thoracodorsal pedicle from its axillary origin to the neurovascular hilum ranged from 10 to 14 cm. With intra-muscular dissection this had been increased by a further 5e10 cm. The majority of segmental flaps were cut parallel to the muscle fasciculi and at micro-dissection had an axial segmental artery running along their length. In this way muscle fibres were not cut but dissected free from each other in a longitudinal fashion. This pattern of dissection should mean that motor end plates were preserved. In all cases the branch of the thoracodorsal nerve followed the vessels within the substance of the muscle, indicating an axial pattern of both blood and nerve supply. In histological sections the diameter of the axial artery ranged from 1.7 mm at its proximal end to 0.6 mm and 0.4 mm distally. Similarly, the segmental nerve diameter ranged from 0.5 mm and 0.49 mm to 0.45 mm along the flap
Figure 7 Haemotoxylin and Eosin (H&E) stained section taken from distal part of same segmental flap. Division of the segmental artery (A) can be seen, in addition to the segmental vein (V) and nerve (N).
1422 length (Table 1). The vessels and nerves were surrounded by connective tissue and ran in the fascia between muscle fibres as distinct neurovascular bundles, (Figures 6 and 7). Vessels, nerves and the surrounding muscle fibres were all seen in transverse section, indicating they ran together along the flap length. At the simulated operation, the transplanted flaps were suitable for the dimensions of the cadaveric face and the donor nerve was long enough to reach the contralateral side. There remained a size mismatch between the diameter of the thoracodorsal nerve and the recipient facial nerve branch. As the facial nerve branches are larger in the lateral areas of the cheek i.e. closer to the parotid gland than to the mouth, we can see that the longer the nerve on the flap, the greater the chance of being able to coapt the flap nerve onto a sizeable branch of the facial nerve. In other words a long pedicle nerve allows more choice in selection of a suitable and sizeable branch of the facial nerve in the non-paralysed side. Studies have suggested that using a larger diameter recipient nerve may correspond to improved physiological outcome.14 It is safe to coapt between two to four nerve grafts using careful technique and nerve stimulation15 however in practice we normally sacrifice up to two branches. The facial artery and vein on the “paralysed” side were connected to the corresponding thoracodorsal artery and vein of the latissimus dorsi flaps without difficulty.
Discussion Our study successfully confirmed the anatomical basis for raising a segmental latissimus dorsi flap with two or more small muscle paddles and a long neurovascular pedicle. In total eleven latissimus dorsi flaps had their pedicle dissected intra-muscularly, confirming three to five segmental neurovascular branches. Histological analysis confirmed the impression that there is a reliable pattern of nerve accompanying the vein and artery distally into the muscle. Clinical experience in four complex cases where facial reanimation using free tissue transfer is required has confirmed the practical application of this concept. Although it remains the senior author’s view that a crossface nerve graft and subsequent free tissue transfer of a smaller muscle remains the gold standard in facial reanimation, in difficult cases where this is not practical, the one-stage micro-dissected latissimus dorsi is a useful and practical alternative.
Primary branching of thoracodorsal pedicle The thoracodorsal pedicle entered latissimus dorsi on its deep surface where it bifurcated into medial and lateral divisions. The nerve divided 2e3 cm proximal to the hilum. These findings were in keeping with those of Tobin and ElMaasarany et al. and laid the foundation for surgically splitting latissimus dorsi into medial and lateral halves for a variety of reconstructive procedures.8,16 Our study indicates the possibility of raising up to four to five small segmental muscle flaps, each of which is chimeric. In reality most facial palsy cases would require one or two such segmental pieces of muscle.
L.D. Ferguson et al.
Segmental (secondary) branching of thoracodorsal pedicle The majority of flaps received an axial segmental artery and nerve along their length. The vessels were seen to run along the deep surface of the muscle. The diameters of a typical segmental artery and nerve on entry into the flap measured 1.7 mm and 1.0 mm (combined diameter for bifurcated nerve) (Table 1). The former is in keeping with Wei et al., who quoted the diameter of the larger thoracodorsal artery to range from 1.6 to 2.7 mm.11 Although the segmental artery and nerve decreased in diameter along the flap length by 40% and 55% respectively, these branches seemed to be of sufficient size to supply the distal end of the flap.
Re-innervation of muscle segments There is some legitimate concern that the micro-dissection needed to lengthen the pedicle might damage the nerve supply of the transplanted segment and weaken its contraction at its new site. In most muscles the motor end plates lie in a single band close to the geometric centre of the muscle, but in latissimus dorsi the muscle fibres are shorter than the length of the fasciculi and the motor end plates are spread over the whole muscle.17 This makes it difficult to predict how many end plates are contained in a given surgical flap and further work is needed to establish this. For the same reason it is not possible to be sure how many end plates are denervated when the nerves are freed by micro-dissection to extend the reach of the pedicle. However, this can be minimised through dissection parallel to the muscle fasciculi. It is also important to remember that after the thoracodorsal nerve is divided at operation the distal axons decay and are replaced by sprouts from the facial nerve branch which is sutured to the stump. These new sprouts are likely to re-innervate some of the surviving end plates in the transplanted segment, but also to stimulate the growth of new ectopic end plates on fibres which are not innervated.18 Because of this mechanism the position of the end plates and transient damage at operation may not have a lasting effect on re-innervation of the graft. This speculation is borne out by our clinical experience which clearly showed recovery of muscle contraction following this dissection.
Conclusion The study showed that the thoracodorsal pedicle is between 10 and 14 cm long and is not long enough to reach easily across the face in a standard dissection. To ensure adequate length it is necessary to raise the flap more distally in the muscle, beyond the origin of the segmental branches. This study clearly shows the anatomical basis for this additional microdissection. It was also noted that segmental arteries and nerves reached the distal end of all flaps examined histologically, implying the potential of a good blood and nerve supply. This has been corroborated with four clinical examples where the thoracodorsal pedicle has been microdissected to give increased length. All of these cases were difficult clinical problems either due to previous oncological
Applied anatomy of the latissimus dorsi free flap surgery meaning vessels for microsurgery were depleted or because a cross-face nerve graft was not possible. Two of the patients had received radiotherapy to the surgical field. The micro-dissected latissimus dorsi is a powerful tool to allow for one-stage facial reanimation in difficult cases. This is because it has a long vascular pedicle to allow recipient vessel access deep into the neck and a long nerve pedicle to allow the nerve to be crossed to a recipient nerve in the contralateral face if required. This study supports the use of the latissimus dorsi muscle where one-stage facial reanimation is required in difficult cases. It is particularly useful in cases where a long neurovascular pedicle is required and a cross-face nerve graft is contraindicated. It has been used successfully to allow facial reanimation even in the post free flap/post radiotherapy patients. Our results are not directly comparable with straightforward cases where a two stage technique is possible. In four very difficult cases the micro-dissected free latissimus dorsi flap has allowed successful facial reanimation and gives hope that reanimation can be offered to more patients as surgical techniques improve.
Conflict of Interest None.
Funding None.
Acknowledgements The authors would like to thank Ashley Conaghan for dissection and Andrew Lockhart for technical laboratory assistance.
References 1. Harrison DH. The treatment of unilateral and bilateral facial palsy using free muscle transfers. Clin Plast Surg 2002;29(4): 539e49. 2. Harii K. Microneurovascular free muscle transplantation for reanimation of facial paralysis. Clin Plast Surg 1979;6(3): 361e75.
1423 3. Bove A, Chiarini S, D’Andrea V, Di Matteo FM, Lanzi G, De Antoni E. Facial nerve palsy: which flap? Microsurgical, anatomical, and functional considerations. Microsurgery 1998; 18:286e9. 4. Mathes SJ, Nahai F. Classification of the vascular anatomy of muscles: experimental and clinical correlation. Plast Reconstr Surg 1981;67(2):177e87. 5. Beer GM, Lang A, Manestar M, Kompatscher P. The bipedicled and bipartite latissimus dorsi free and perforator flap: an anatomic study. Plast Reconstr Surg 2006;118(5):1162e70. 6. Takushima A, Harii K, Asato H, Momosawa A, Okazaki M. One-stage reconstruction of facial paralysis associated with skin/soft tissue defects using latissimus dorsi compound flap. J Plast Reconstruct Aesthet Surg 2006;59(5):465e73. 7. Bartlett SP, May JW, Yaremchuk MJ. The latissimus dorsi muscle: a fresh cadaver study of the primary neurovascular pedicle. Plast Reconstr Surg 1981;67(5):631e6. 8. Tobin GR, Schusterman M, Peterson GH, Nichols G, Bland KI. The intramuscular neurovascular anatomy of the latissimus dorsi muscle: the basis for splitting the flap. Plast Reconstr Surg 1981;67(5):637e41. 9. Zhao L, Miao H, Wang W, Zhang D. The anatomy of the segmental latissimus dorsi flap for reconstruction of facial paralysis. Surg Radiol Anat 1993;15:239e43. 10. Wong MTC, Lim AY, Coninck CD, Kumar PV. Functional units within the latissimus dorsi muscle based on Sihler technique. Ann Plast Surg 2007;59:152e5. 11. Wei W, Zuoliang Q, Xiaoxi L, et al. Free split and segmental latissimus dorsi muscle transfer in one stage for facial reanimation. Plast Reconstr Surg 1999;103(2):473e80. 12. Watanabe K, Kiyokawa K, Rikimaru H, Koga N, Yamaki K, Saga T. Anatomical study of latissimus dorsi musculocutaneous flap vascular distribution. J Plast Reconstruct Aesthet Surg 2010;63(7):1091e8. 13. Bancroft JD, Stevens A. Theory and practice of histological techniques. 2nd ed. Edinburgh: Churchill Livingstone; 1982. 14. MacQuillan AH, Grobbelaar AO. Functional muscle transfer and the variance of reinnervating axonal load: part I. The facial nerve. Plast Reconstr Surg 2008;121(5):1570e7. 15. Terzis JK, Konofaos P. Nerve transfers in facial palsy. Facial Plast Surg 2008;24(2):177e93. 16. EI-Maasarany SH, Sharaf E, Moustafa F, Borhan A, AbdelFattah A, Hamza A. Anatomical basis of latissimus dorsi myocutaneous flap: clinical applications. Surg Radiol Anat 1989; 11:197e203. 17. Snobl D, Binaghi LE, Zenker W. Microarchitecture and innervation of the human latissimus dorsi muscle. J Reconstr Microsurg 1998;14(3):171e7. 18. Kernell D. The motoneurone and its muscle fibres (Monographs of the physiological society). 1st ed. Oxford: Oxford University Press; 2006.
Applied anatomy of the latissimus dorsi free flap for refinement in one-stage facial reanimation* L.D. Ferguson a,*, T. Paterson a, F. Ramsay a, K. Arrol a, J. Dabernig b, J. Shaw-Dunn a, S. Morley b a
Department of Human Anatomy, Thomson Building, University of Glasgow, Glasgow G12 8QQ, UK Canniesburn Plastic Surgery Unit, Glasgow Royal Infirmary, Jubilee Building, 84 Castle Street, Glasgow G4 0SF, UK
b
Received 4 December 2010; accepted 7 June 2011
KEYWORDS Latissimus dorsi flap; Segmental; Facial reanimation; Thoracodorsal pedicle
Summary Background: The face can be reanimated after long-term paralysis by free microneurovascular tissue transfer. Flaps from gracilis and pectoralis minor usually require a twostage procedure with a cross-face nerve graft. Latissimus dorsi has a much longer muscular nerve, the thoracodorsal nerve, which could avoid the need for a second cross-face nerve graft. Our hypothesis is that the neurovascular pedicles of small segments of latissimus dorsi would be long enough to reach the opposite side of the face and to provide a reliable blood and nerve supply to the flaps. Method: To test this hypothesis the thoracodorsal pedicle and its primary branches were dissected in eleven embalmed cadavers. The segmental vessels and nerves were then traced in a series of simulated flaps approximately 8e10 cm 2e3 cm by micro-dissection, tissue clearing and histology. Results: The thoracodorsal pedicle is 10e14 cm long to where it enters the muscle, and with intra-muscular dissection small chimeric muscle segments 8e10 cm 2e3 cm can be raised with a clearly defined neurovascular supply. Using micro-dissection the neurovascular pedicle can be lengthened to reach across the face. Segmental arteries and nerves extended to the distal end of all the flaps examined. Artery, vein and nerve run together and are of substantial diameter. Conclusion: Small muscle segments of latissimus dorsi can be raised on long neurovascular pedicles. The vessels and nerves are substantial and the likelihood of surgical complications such as flap necrosis and functional disuse on transplantation appear low. Although in our
* Presentations: Research related to this article has been presented by the corresponding author LD Ferguson at the Anatomical Society of Great Britain and Ireland (ASGBI) Winter Meeting January 2009 and by T Paterson at the British Association of Clinical Anatomists (BACA) Winter Meeting December 2008, at which it was awarded the Conrad Lewin prize. * Corresponding author. E-mail address: [email protected] (L.D. Ferguson).
1748-6815/$ - see front matter ª 2011 British Association of Plastic, Reconstructive and Aesthetic Surgeons. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.bjps.2011.06.013
1418
L.D. Ferguson et al. opinion the use of cross-face nerve grafts and transfer of smaller muscle flaps remains the gold standard in facial reanimation in straightforward cases, the micro-dissected latissimus dorsi flap is a useful option in complex cases of facial reanimation. Clinical Application: Facial reanimation using micro-dissected segments of latissimus dorsi has been performed in four complex cases of facial paralysis. ª 2011 British Association of Plastic, Reconstructive and Aesthetic Surgeons. Published by Elsevier Ltd. All rights reserved.
Introduction Facial paralysis is disfiguring, with emotional expression, chewing, speech, and eye closure affected and asymmetry of the face at rest. Surgical reanimation techniques for irreversible facial paralysis mostly focus on attempts to restore voluntary facial movement and emotional expression i.e. to recreate the smile. Irreversible can be taken to mean long standing facial paralysis which is unlikely to recover either due to the nature of the injury or the time elapsed. Several potential donor muscles have been used, the commonest muscles being the gracilis and pectoralis minor flaps, usually implanted as a second stage following a cross-face nerve graft.1,2 In complex cases or where a single stage procedure is desired a larger muscle with a longer pedicle may be used and the most frequently used in this circumstance is latissimus dorsi. This is due to its long neurovascular pedicle and low donor site morbidity.3 The latissimus dorsi is particularly useful where a long pedicle is required due to vessel depletion in the neck or where a cross-face nerve graft cannot be used because a one-stage reconstruction is required. Latissimus dorsi is a type V muscle in the classification of Mathes and Nahai (1981) with one dominant vascular pedicle from the thoracodorsal artery at its insertion, and several secondary perforator pedicles arising from the 9th, 10th, and 11th terminal intercostal and 1st, 2nd, and 3rd lumbar arteries, at its origin.4 Beer (2006) has argued the merits of a “bipedicled” flap based on this dual supply5 but the majority of latissimus dorsi free flaps are still designed solely on the dominant thoracodorsal pedicle.6 In its entirety the muscle is far too large to be transplanted to the face as a whole and so a basis has been sought for splitting the muscle into viable smaller segments with a long pedicle. Initial studies by Bartlett (1981) and Tobin (1981) revealed primary bifurcation of the thoracodorsal vessels into medial and lateral branches.7,8 The key finding however was the subsequent discovery of 4e6 segmental (secondary) arteries and nerves by Zhao et al. (1993) and Wong et al. (2007).9,10 These findings suggested that latissimus dorsi could be split into 4e6 smaller segmental muscle flaps as independent, functional units for transplantation to the face.11 A recent paper has looked at the latissimus dorsi pedicle in relation to its three dimensional blood supply both to the muscle and the overlying skin, showing a consistent branching pattern in three separate vascular territories with choke vessels joining these areas.12 For segmental muscle flaps to be successful as a one-stage facial reanimation technique the neurovascular pedicle of latissimus dorsi must be long enough to reach across the face, and the segmental flaps must have an adequate blood and nerve supply throughout their length. Several authors have
used the latissimus dorsi muscle in this way with success. We wished to determine more precisely the anatomic basis of the muscle as the pedicle splits and passes from proximal to distal. In particular we wished to determine for how far along its length is the segmental neurovascular supply of the latissimus dorsi muscle predictable. To this end several gross dissections of muscle were made; histology was used to determine the microscopic arrangement of vessels and nerve and simulated reanimation operations with segmental flaps of latissimus dorsi were undertaken. The knowledge gained by these studies has been applied in four cases of complex facial reanimation. These cases demonstrated the need either for a long vascular pedicle required because of previous surgery or a long nerve to reach the contralateral side of the face.
Materials and methods Tissue used in this study was obtained from eleven cadavers donated to the Department of Anatomy, University of Glasgow. These had been prepared by common carotid artery injection of embalming fluid followed after 48 h by a threepart dilution of ammonia and, finally, a mixture of India ink and latex rubber. Human tissue is held under the provisions of the Anatomy Act (1984) as revised in the Human Tissue (Scotland) Act 2006, which includes ethical approval.
Gross- and micro-dissection A mid-axillary incision was used to locate the distal third of the axillary artery from which the subscapular artery arose. The constituents of the neurovascular pedicle, the thoracodorsal artery, vein and nerve were identified and traced to their entry into latissimus dorsi at the neurovascular hilum. In each of eleven latissimus dorsi muscles, the senior author fashioned four to five segments approximately 8e10 cm 2e3 cm, based on the visible branches of the thoracodorsal artery and nerve, as at operation. These pieces of muscle were dissected distal to the point of insertion of the neurovascular pedicle meaning muscle flap segments were raised from approximately 12e14 cm to 20e24 cm from the origin of the muscle in the axilla (Figures 1 and 2). The flaps were dissected with the aid of a dissecting microscope to trace the segmental artery along the full length of each flap. The main focus of the study was on raising lateral segmental flaps of muscle placed distally, as would be used in reanimation surgery.
Tissue clearing To trace the smaller branches of the arteries the muscle was cleared. In this technique, the muscle is fully
Applied anatomy of the latissimus dorsi free flap
1419
Simulated facial reanimation
Figure 1 Deep surface of latissimus dorsi muscle surgically split into 4 segmental flaps based on its segmental branches (white asterisks). Segments 1 and 2 are the most lateral and readily exposed at operation. The thoracodorsal pedicle, along with its lateral and medial branches, is also visible.
dehydrated in alcohol then immersed in methyl salicylate which makes the tissue translucent.13 In the final result, the arteries filled with Indian ink appear jet black on a golden yellow background in which the main tissue elements are still recognisable.
Histology Full thickness transverse sections were taken at 4 cm intervals along the segmental muscle flaps and processed for routine histology. Sections were taken at specified points on the pathway of the neurovascular pedicle from proximal to distal with particular emphasis on the pedicle arrangement distally. These flaps would mimic the piece of muscle which would be used in facial reanimation using a pedicle sufficiently long to reach the contralateral cheek. Arterial and nerve diameters in the sections were measured with a Zeiss AxioSkop2 Mot Plus microscope with Axiocam for digital capture and Zeiss AxioVision software version 4.7, (Carl Zeiss MicroImaging GmbH, Gottingen, Germany).
At the simulated operation, two segmental flaps of latissimus dorsi measuring 8 cm in length and 3 cm in width were raised on extended pedicles which had been lengthened with intra-muscular dissection. They were then attached into position medially at the modiolus and angled laterally onto the temporal fascia. The thoracodorsal nerve was passed through the buccal cavity of the cheek with a pair of forceps into the mouth. From the mouth, the nerve was tunnelled through the upper lip and fed into the contralateral side of the face. Here the thoracodorsal nerve met the pre-dissected branches of the facial nerve. This mock procedure confirmed that the nerve, once appropriately dissected, is long enough to pass across the face for coaptation to the contralateral facial nerve, exposed through a pre-auricular incision.
Clinical examples Since the anatomical side of the study was completed, a segmental latissimus dorsi flap has been used in complex facial reanimation in four cases. In each case the anatomical knowledge gained from the cadaver studies was used to expedite micro-dissection of precisely formed segments of the latissimus dorsi muscle, on a long pedicle. Flaps were raised using a lateral incision in the axilla with the patient in the supine position.
Case one A 44 year old female presented to the facial palsy clinic with a dense left facial paralysis secondary to radical excisional surgery followed by chemoradiotherapy to treat a recurrent parotid adenocarcinoma. She already had a free flap (myocutaneous rectus flap) on her left cheek and the temporalis muscle had been excised along with the facial nerve. It was decided that a one-stage free tissue transfer using a latissimus dorsi muscle would give the best chance of both an improved static position and restoration of smile. Due to her previous surgery this would have to be motored onto the contralateral facial nerve and recipient vessels would be located deep in the neck, meaning a long vascular pedicle was required. Two chimeric muscle paddles of 8 cm 3 cm were raised with a 15 cm length of nerve, and inset into the left cheek in front of the pre-existing free flap. The thoracodorsal nerve was coapted onto a suitable branch of the facial nerve on the contralateral side, exposed through a pre-auricular incision. At 18 months follow up this has given an improved static position where the position of the modiolus has been lifted 9 mm to leave it symmetrical with the normal side. Dynamic correction of the palsy which is spontaneous with the patient’s smile, has also been achieved with 5 mm of modiolus excursion, although the smile is not fully symmetrical.
Case two
Figure 2 Latissimus dorsi has been surgically split into 5 segmental flaps in this case based upon the 5 segmental vessels and nerves (A to E).
A 51 year old female presented to the facial palsy clinic with a dense left facial paralysis secondary to radical excisional surgery and radiotherapy to treat a maxillary carcinoma. She had been left with a mandibular defect due
1420
L.D. Ferguson et al. of 10 cm 3 cm was raised on a micro-dissected intramuscular pedicle 10 cm distal to the main bulk of the flap. The nerve was crossed to the contralateral facial nerve and coapted to a suitable buccal branch. At 6 months follow up the procedure has resulted in an improved static position with 5 mm of modiolus elevation and 3 mm of modiolus excursion when smiling.
Case three
Figure 3 Deep surface of latissimus dorsi still attached to the body. The lateral and medial branches of the thoracodorsal pedicle can be seen at the neurovascular hilum. From these arise the segmental vessels and nerves upon which the muscle flaps are based (Figure 1). Secondary perforators can also be seen at its origin.
to severe osteo-radionecrosis. The original excisional surgeon required to place a muscle free flap to correct the mandibular soft tissue defect. It was decided therefore to raise a chimeric latissimus dorsi, with the main bulk of the flap being used to correct this defect but with a segmental muscle paddle being used for facial reanimation. A segment
A 32 year old female presented to the facial palsy clinic having had previous facial reanimation surgery at another unit with the insertion of a free rectus flap performed in another unit. Her initial palsy was related to excision of a vestibular schwannoma. The muscle insertion had dehisced both proximally and distally meaning that the muscular contraction of the muscle was producing a bunched up appearance. One attempt to reposition the existing muscle was unsuccessful and following this surgery there was only a flicker of muscle movement. It was decided to remove the existing muscle and replace it with a second muscle transfer. Due to her previous surgery we knew a long vascular pedicle would be required so a segmental latissimus dorsi flap was raised. As the crossface nerve graft was not seen to be working strongly the thoracodorsal nerve was tunnelled to the contralateral side to provide a fresh neural input. The full length of the pedicle was required as the lingual artery was used as a recipient vessel. At 13 months follow up the operation has successfully improved her static position by a 6 mm lift at the modiolus and she now has spontaneous movement of the modiolus of 5 mm when smiling.
Case four A 44 year old male with bilateral facial paralysis secondary to acoustic neuroma surgery presented to clinic complaining of
Figure 4 Another latissimus dorsi flap still attached to the body by the thoracodorsal (TD) pedicle. In this case, the pedicle divides into 5 segmental branches upon which the flaps in Figure 2 are based. A is the lateral branch.
Figure 5 Latissimus dorsi after clearing. The arteries injected with Indian Ink appear black. There are three substantial arteries which spread out parallel to the radiating muscle fasciculi. The lateral branch gives two subdivisions (A & B) giving in all four segmental arteries AeD. Anastomoses can be traced with secondary perforators to the right of the figure.
Applied anatomy of the latissimus dorsi free flap Table 1
1421
Segmental artery and nerve diameters at proximal, middle, and distal aspects of a typical segmental flap.
Distance along muscle segment:
Diameter of segmental artery (mm):
Diameter of segmental nerve (mm):
Proximal (0 cm) Middle (4 cm) Distal (8 cm)
1.74 0.80 0.62 and 0.41
0.53 and 0.49 0.34 and 0.28 0.45
asymmetry and complete lack of facial movement on the right side. Many years previously he had had an attempt at facial reanimation using a free gracilis muscle transfer which had not been successful in re-animating his face. He had bilateral fascia lata static supports added at an earlier procedure. The patient received a segmentally split latissimus dorsi flap motored onto the nerve to masseter as a one-stage procedure. The pedicle was dissected into the muscle to allow a length of pedicle to be harvested sufficient to reach deep into the neck to obtain suitable vessels for micro-anastamosis which was required as the facial vessels in the cheek had been utilised previously for microsurgery. Although the superficial temporal vessels could potentially have been used as recipient vessels due to the patient’s previous procedures it was not felt that these would have been reliable. At 6 months follow up the combination of fascia lata supports and free tissue transfer have successfully improved his static position by a 20 mm lift at the modiolus and he now has movement of the modiolus of 7 mm when smiling.
Results The latissimus dorsi is a large triangular superficial muscle of the back with a narrow insertion into the intertubercular groove of the humerus. The lower fibres are near vertical, the middle oblique, and the upper fibres almost horizontal. The lateral border is readily exposed at dissection and at operation. The thoracodorsal artery divides as it approaches the neurovascular hilum on the deep surface of
Figure 6 Miller and Van Gieson stained section taken from proximal part of a typical segmental flap. The segmental artery (A) with two accompanying segmental veins (V) and nerves (N) are seen surrounded by connective tissue in a neurovascular bundle.
the muscle. The usual pattern is two divisions, lateral and medial (Figure 3), but occasionally there are three. Within the muscle, segmental branches spread out parallel to the radiating muscle fasciculi. These branches vary in number and diameter but in our series of eleven cadaver dissections there were never fewer than three substantial branches and often four or five (Figures 1, 4, and 5). The lateral branch follows the lateral border of the muscle at a distance of 1e2 cm and often divides into two major subdivisions (Figure 5). There are large anastomoses with the secondary perforators (Figure 5). The thoracodorsal nerve divides proximal to the artery. The branches usually follow the arteries in their radiating course, but occasionally the main trunks may diverge. The total length of the thoracodorsal pedicle from its axillary origin to the neurovascular hilum ranged from 10 to 14 cm. With intra-muscular dissection this had been increased by a further 5e10 cm. The majority of segmental flaps were cut parallel to the muscle fasciculi and at micro-dissection had an axial segmental artery running along their length. In this way muscle fibres were not cut but dissected free from each other in a longitudinal fashion. This pattern of dissection should mean that motor end plates were preserved. In all cases the branch of the thoracodorsal nerve followed the vessels within the substance of the muscle, indicating an axial pattern of both blood and nerve supply. In histological sections the diameter of the axial artery ranged from 1.7 mm at its proximal end to 0.6 mm and 0.4 mm distally. Similarly, the segmental nerve diameter ranged from 0.5 mm and 0.49 mm to 0.45 mm along the flap
Figure 7 Haemotoxylin and Eosin (H&E) stained section taken from distal part of same segmental flap. Division of the segmental artery (A) can be seen, in addition to the segmental vein (V) and nerve (N).
1422 length (Table 1). The vessels and nerves were surrounded by connective tissue and ran in the fascia between muscle fibres as distinct neurovascular bundles, (Figures 6 and 7). Vessels, nerves and the surrounding muscle fibres were all seen in transverse section, indicating they ran together along the flap length. At the simulated operation, the transplanted flaps were suitable for the dimensions of the cadaveric face and the donor nerve was long enough to reach the contralateral side. There remained a size mismatch between the diameter of the thoracodorsal nerve and the recipient facial nerve branch. As the facial nerve branches are larger in the lateral areas of the cheek i.e. closer to the parotid gland than to the mouth, we can see that the longer the nerve on the flap, the greater the chance of being able to coapt the flap nerve onto a sizeable branch of the facial nerve. In other words a long pedicle nerve allows more choice in selection of a suitable and sizeable branch of the facial nerve in the non-paralysed side. Studies have suggested that using a larger diameter recipient nerve may correspond to improved physiological outcome.14 It is safe to coapt between two to four nerve grafts using careful technique and nerve stimulation15 however in practice we normally sacrifice up to two branches. The facial artery and vein on the “paralysed” side were connected to the corresponding thoracodorsal artery and vein of the latissimus dorsi flaps without difficulty.
Discussion Our study successfully confirmed the anatomical basis for raising a segmental latissimus dorsi flap with two or more small muscle paddles and a long neurovascular pedicle. In total eleven latissimus dorsi flaps had their pedicle dissected intra-muscularly, confirming three to five segmental neurovascular branches. Histological analysis confirmed the impression that there is a reliable pattern of nerve accompanying the vein and artery distally into the muscle. Clinical experience in four complex cases where facial reanimation using free tissue transfer is required has confirmed the practical application of this concept. Although it remains the senior author’s view that a crossface nerve graft and subsequent free tissue transfer of a smaller muscle remains the gold standard in facial reanimation, in difficult cases where this is not practical, the one-stage micro-dissected latissimus dorsi is a useful and practical alternative.
Primary branching of thoracodorsal pedicle The thoracodorsal pedicle entered latissimus dorsi on its deep surface where it bifurcated into medial and lateral divisions. The nerve divided 2e3 cm proximal to the hilum. These findings were in keeping with those of Tobin and ElMaasarany et al. and laid the foundation for surgically splitting latissimus dorsi into medial and lateral halves for a variety of reconstructive procedures.8,16 Our study indicates the possibility of raising up to four to five small segmental muscle flaps, each of which is chimeric. In reality most facial palsy cases would require one or two such segmental pieces of muscle.
L.D. Ferguson et al.
Segmental (secondary) branching of thoracodorsal pedicle The majority of flaps received an axial segmental artery and nerve along their length. The vessels were seen to run along the deep surface of the muscle. The diameters of a typical segmental artery and nerve on entry into the flap measured 1.7 mm and 1.0 mm (combined diameter for bifurcated nerve) (Table 1). The former is in keeping with Wei et al., who quoted the diameter of the larger thoracodorsal artery to range from 1.6 to 2.7 mm.11 Although the segmental artery and nerve decreased in diameter along the flap length by 40% and 55% respectively, these branches seemed to be of sufficient size to supply the distal end of the flap.
Re-innervation of muscle segments There is some legitimate concern that the micro-dissection needed to lengthen the pedicle might damage the nerve supply of the transplanted segment and weaken its contraction at its new site. In most muscles the motor end plates lie in a single band close to the geometric centre of the muscle, but in latissimus dorsi the muscle fibres are shorter than the length of the fasciculi and the motor end plates are spread over the whole muscle.17 This makes it difficult to predict how many end plates are contained in a given surgical flap and further work is needed to establish this. For the same reason it is not possible to be sure how many end plates are denervated when the nerves are freed by micro-dissection to extend the reach of the pedicle. However, this can be minimised through dissection parallel to the muscle fasciculi. It is also important to remember that after the thoracodorsal nerve is divided at operation the distal axons decay and are replaced by sprouts from the facial nerve branch which is sutured to the stump. These new sprouts are likely to re-innervate some of the surviving end plates in the transplanted segment, but also to stimulate the growth of new ectopic end plates on fibres which are not innervated.18 Because of this mechanism the position of the end plates and transient damage at operation may not have a lasting effect on re-innervation of the graft. This speculation is borne out by our clinical experience which clearly showed recovery of muscle contraction following this dissection.
Conclusion The study showed that the thoracodorsal pedicle is between 10 and 14 cm long and is not long enough to reach easily across the face in a standard dissection. To ensure adequate length it is necessary to raise the flap more distally in the muscle, beyond the origin of the segmental branches. This study clearly shows the anatomical basis for this additional microdissection. It was also noted that segmental arteries and nerves reached the distal end of all flaps examined histologically, implying the potential of a good blood and nerve supply. This has been corroborated with four clinical examples where the thoracodorsal pedicle has been microdissected to give increased length. All of these cases were difficult clinical problems either due to previous oncological
Applied anatomy of the latissimus dorsi free flap surgery meaning vessels for microsurgery were depleted or because a cross-face nerve graft was not possible. Two of the patients had received radiotherapy to the surgical field. The micro-dissected latissimus dorsi is a powerful tool to allow for one-stage facial reanimation in difficult cases. This is because it has a long vascular pedicle to allow recipient vessel access deep into the neck and a long nerve pedicle to allow the nerve to be crossed to a recipient nerve in the contralateral face if required. This study supports the use of the latissimus dorsi muscle where one-stage facial reanimation is required in difficult cases. It is particularly useful in cases where a long neurovascular pedicle is required and a cross-face nerve graft is contraindicated. It has been used successfully to allow facial reanimation even in the post free flap/post radiotherapy patients. Our results are not directly comparable with straightforward cases where a two stage technique is possible. In four very difficult cases the micro-dissected free latissimus dorsi flap has allowed successful facial reanimation and gives hope that reanimation can be offered to more patients as surgical techniques improve.
Conflict of Interest None.
Funding None.
Acknowledgements The authors would like to thank Ashley Conaghan for dissection and Andrew Lockhart for technical laboratory assistance.
References 1. Harrison DH. The treatment of unilateral and bilateral facial palsy using free muscle transfers. Clin Plast Surg 2002;29(4): 539e49. 2. Harii K. Microneurovascular free muscle transplantation for reanimation of facial paralysis. Clin Plast Surg 1979;6(3): 361e75.
1423 3. Bove A, Chiarini S, D’Andrea V, Di Matteo FM, Lanzi G, De Antoni E. Facial nerve palsy: which flap? Microsurgical, anatomical, and functional considerations. Microsurgery 1998; 18:286e9. 4. Mathes SJ, Nahai F. Classification of the vascular anatomy of muscles: experimental and clinical correlation. Plast Reconstr Surg 1981;67(2):177e87. 5. Beer GM, Lang A, Manestar M, Kompatscher P. The bipedicled and bipartite latissimus dorsi free and perforator flap: an anatomic study. Plast Reconstr Surg 2006;118(5):1162e70. 6. Takushima A, Harii K, Asato H, Momosawa A, Okazaki M. One-stage reconstruction of facial paralysis associated with skin/soft tissue defects using latissimus dorsi compound flap. J Plast Reconstruct Aesthet Surg 2006;59(5):465e73. 7. Bartlett SP, May JW, Yaremchuk MJ. The latissimus dorsi muscle: a fresh cadaver study of the primary neurovascular pedicle. Plast Reconstr Surg 1981;67(5):631e6. 8. Tobin GR, Schusterman M, Peterson GH, Nichols G, Bland KI. The intramuscular neurovascular anatomy of the latissimus dorsi muscle: the basis for splitting the flap. Plast Reconstr Surg 1981;67(5):637e41. 9. Zhao L, Miao H, Wang W, Zhang D. The anatomy of the segmental latissimus dorsi flap for reconstruction of facial paralysis. Surg Radiol Anat 1993;15:239e43. 10. Wong MTC, Lim AY, Coninck CD, Kumar PV. Functional units within the latissimus dorsi muscle based on Sihler technique. Ann Plast Surg 2007;59:152e5. 11. Wei W, Zuoliang Q, Xiaoxi L, et al. Free split and segmental latissimus dorsi muscle transfer in one stage for facial reanimation. Plast Reconstr Surg 1999;103(2):473e80. 12. Watanabe K, Kiyokawa K, Rikimaru H, Koga N, Yamaki K, Saga T. Anatomical study of latissimus dorsi musculocutaneous flap vascular distribution. J Plast Reconstruct Aesthet Surg 2010;63(7):1091e8. 13. Bancroft JD, Stevens A. Theory and practice of histological techniques. 2nd ed. Edinburgh: Churchill Livingstone; 1982. 14. MacQuillan AH, Grobbelaar AO. Functional muscle transfer and the variance of reinnervating axonal load: part I. The facial nerve. Plast Reconstr Surg 2008;121(5):1570e7. 15. Terzis JK, Konofaos P. Nerve transfers in facial palsy. Facial Plast Surg 2008;24(2):177e93. 16. EI-Maasarany SH, Sharaf E, Moustafa F, Borhan A, AbdelFattah A, Hamza A. Anatomical basis of latissimus dorsi myocutaneous flap: clinical applications. Surg Radiol Anat 1989; 11:197e203. 17. Snobl D, Binaghi LE, Zenker W. Microarchitecture and innervation of the human latissimus dorsi muscle. J Reconstr Microsurg 1998;14(3):171e7. 18. Kernell D. The motoneurone and its muscle fibres (Monographs of the physiological society). 1st ed. Oxford: Oxford University Press; 2006.