enhancing the outcome of free latissimus dorsi muscle flap

ENHANCING THE OUTCOME OF FREE LATISSIMUS DORSIMUSCLE FLAP RECONSTRUCTION OF SCALP DEFECTS

Joan E. Lipa, MD, MSc, FRCS(C), Charles E. Butler, MD, FACS

Department of Plastic Surgery, The University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Blvd.,

Unit 443, Houston, TX 77030-4009. E-mail: [email protected]

Accepted 22 May 2003

Published online 23 September 2003 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/hed.10338

Abstract: Background. Reconstruction of scalp and calvarial

defects after tumor ablation frequently requires prosthetic

cranioplasty and cutaneous coverage. Furthermore, patients

often have advanced disease and receive perioperative radio-

therapy. We evaluated the complications of scalp reconstruction

with a free latissimus dorsi muscle flap in this setting.

Methods. The complications and the oncologic and aesthetic

outcomes of six consecutive scalp reconstructions with a free

latissimus dorsi muscle flap and skin graft in five patients with

advanced cancer were retrospectively evaluated. Patient, tumor,

defect, reconstructive, and other treatment characteristics were

reviewed. Reconstructive and perioperative techniques in-

tended to improve flap survival and aesthetic outcome and

reduce complications in these patients.

Results. All patients (52–76 years old) had recurrent tumors

(sarcoma, melanoma, or squamous cell carcinoma) and

received postoperative radiotherapy. The mean scalp defect

size was 367 cm2, and partial-thickness or full-thickness

calvarial resection was required in all six cases. No vein grafts

were needed. The mean follow-up period and disease-free

survival time were 18 and 13 months, respectively. Three

patients died of their disease, and two survived disease free.

There were no flap failures or dehiscences. Complications

consisted of donor site seroma in two patients; partial skin graft

loss in one patient; and radiation burns to the flap, face, and ears

in one patient. Scalp contour and aesthetic outcome were very

good in all cases except for the one case with radiation burns.

Conclusions. Good outcomes were achieved using a free

latissimus dorsi muscle flap with a skin graft for flap recon-

struction in elderly patients with advanced recurrent cancers

who received perioperative radiotherapy. Several technical

aspects of the reconstruction technique intended to enhance

the functional and aesthetic outcome and/or reduce complica-

tions were believed to have contributed to the good results.

B 2004 Wiley Periodicals, Inc. Head Neck 26: 46–53, 2004

Keywords: surgical flaps; latissimus dorsi muscle; scalp;

calvarium; microsurgery; squamous cell carcinoma; sarcoma;

skin graft

Reconstruction of scalp defects after oncologic

resection commonly requires free tissue transfer.

Although local flaps can provide single-stage scalp

reconstruction with good aesthetic results, the size

of the defect and previous reconstructive proce-

dures often preclude the use of local flaps. Further-

more, patients are often elderly, have advanced

malignancies, and receive adjuvant radiotherapy.

In addition, the scalp lesions might have recurred

after previous resection and reconstruction, or the

lesions might be ulcerated and contaminated with

bacteria and the remaining calvarial bone widely

exposed and devoid of periosteum.

Since the inception of free tissue transfer, the

free latissimus dorsi muscle flap has been

recognized as an excellent option for scalp

reconstruction because of its large surface area,

long vascular pedicle, and provision of reliable,

well-vascularized tissue.1,2 However, the revision

Correspondence to: C. E. Butler

Presented in part at the 17th Annual Meeting of the American Society forReconstructive Microsurgery, Cancun, Mexico, January 12, 2002.

B 2004 Wiley Periodicals, Inc.

HEAD & NECK January 200446 Technical Refinements In Scalp Reconstruction

rate with free flaps is high—reported as 20% to

32%.3,4 Because many cancer patients requiring

large resections of scalp and calvarium have a poor

prognosis,5 a single-stage surgical procedure—

without the need for revisions—is desirable to

improve the quality of remaining life in these

patients and to allow for undelayed delivery of

postoperative radiotherapy.

Significant wound healing complications occur

in approximately one third of cancer patients

undergoing scalp reconstruction after tumor abla-

tion even when the entire flap survives.4 Potential

complications and adverse outcomes include par-

tial flap necrosis, the need for vein grafts because

of long pedicle length requirement, wound dehis-

cence with bone or cranioplasty exposure, donor

site seromas, and contour irregularities such as

depression of the scalp-flap junction and excessive

bulk at the muscle origin. Methods to reduce flap

and wound healing complications and enhance

aesthetic results would improve the outcome of

scalp reconstruction.

The purpose of this study was to evaluate the

complications and outcomes of scalp reconstruc-

tion using the free latissimus dorsi muscle flap in

patients with advanced malignancies. The tech-

nical aspects of the reconstruction technique that

were used to reduce complications and improve

functional and aesthetic outcome are described

in detail.

PATIENTS AND METHODS

All free latissimus dorsi muscle flap reconstruc-

tions of the scalp performed by the senior author

(CEB) at The University of Texas M. D. Anderson

Cancer Center over a 2-year period (June 30, 1999,

to July 1, 2001) were retrospectively reviewed.

Patients were identified by query of a comprehen-

sive departmental database, and the charts of all

identified patients were reviewed. Patient, tumor,

and defect characteristics, operative details, use

and dose of perioperative radiotherapy, complica-

tions, and outcomes were analyzed.

RESULTS

Six latissimus dorsi muscle flaps were used for

scalp reconstruction in five patients (all men;

mean age, 64.8 years; age range, 52–76 years)

after resection of sarcoma (four flaps), melanoma

(one flap), or squamous cell carcinoma (one flap)

(Table 1). All tumors were local recurrences, and in

all cases the patient had previously undergone a

scalp flap and/or a skin graft reconstruction after

resection. All patients received radiotherapy

(mean dose, 62.4 Gy) postoperatively, and the

patient who received two latissimus dorsi flaps

also had radiotherapy after the second surgery.

The mean scalp defect area was 367 cm2

(range, 195–900 cm2). All six tumor resections

required calvarial bone excision: full-thickness

(mean size, 67 cm2; range, 20–99 cm2) in three

cases and outer table only in three cases (mean

size, 43 cm2; range, 20–90 cm2). The three full-

thickness calvarial defects were reconstructed

with titanium mesh in two cases and methylme-

thacrylate in one case. Resection of the dura was

required in one case (patient 4), which involved

resection of the cerebral cortex; reconstruction of

Table 1. Summary of tumor and defect characteristics and outcomes.

Patient

Tumor

histology*

Calvarial

defect, cm2Scalp

defect, cm2Follow-up,

mo Outcome Complications

1 Angiosarcoma 20 (PT) 900 8 Died, 8 months postop,

metastasis (lungs)

Partial skin graft loss;

severe radiation burns

2a Unclassified sarcoma 20 (FT) 288 27 Local recurrence (scalp),

7 mo postop

None

2b Unclassified sarcoma 30 (PT) 400 20 Alive, NED, 27 mo

postop from primary resection

Donor site seroma

3 Squamous cell carcinoma 90 (PT) 195 26 Alive, NED, 26 mo postop None

4 Angiosarcoma 99y (FT) 210 14 Died, 14 mo postop,

local recurrence (brain)

None

5 Melanoma 81 (FT) 210 18 Died, 18 mo postop,

local recurrence (brain)

and metastasis (lungs and bone)

Donor site seroma

*All were recurrent tumors of the scalp.yIncluded dura and cerebral cortex resection.Abbreviations: FT, full thickness; NED, no evidence of disease; postop, postoperatively; PT, partial thickness.

Technical Refinements In Scalp Reconstruction HEAD & NECK January 2004 47

the dura was performed with allogeneic dura.

Recipient vessels used for the free flaps were the

superficial temporal artery and vein (both end-to-

end) in five cases and the external carotid artery

and internal jugular vein (both end-to-side) in one

case. No vein grafts were required.

No complete or partial flap losses were encoun-

tered. Complications occurred in three patients. In

one patient who underwent a near-total scalp

reconstruction, 7% of the skin graft over the distal

flap area did not survive. Fortunately, the distal

aspect of this large muscle flap was purposely

oriented over the unresected portion of the right

temporalis muscle (rather than bone) to facilitate

treatment if vascular insufficiency of the flap

occurred in this location (Figure 1). This area

likely would have reepithelialized after debride-

ment and dressing changes; however, a skin graft

was placed to allow prompt initiation of radio-

therapy. The patient had significant radiation

burns to his face, ear, neck, chest, and flap site and

refused further treatment because of the develop-

ment of rapidly progressive metastatic disease.

The other two patients with complications had a

seroma develop in the flap donor site; both

FIGURE 1. (A) Near-total scalp defect after radical excision of an angiosarcoma in a 52-year-old man. (B) The latissimus dorsi muscle

was inset into the defect after microsurgical anastomosis to the left superficial temporal vessels. The distal (least reliable) aspect of the

flap was intentionally oriented over the unresected portion of right temporalis muscle to facilitate treatment if this portion of the flapdeveloped vascular insufficiency.

Table 2. Technical considerations to improve outcome.

Purpose Techniques

I. Improve flap survival. Wide undermining and scoring of galea in pedicle tunnel. Orientation of submuscular drainage catheter parallel to flap

vascular axis. Diligent postoperative patient positioning to prevent com-

pression of flap and pedicle

II. Reduce pedicle length requirements (and need for vein

grafts). Use of superficial temporal recipient vessels when possible. Intramuscular pedicle dissection. Incision/excision of scalp tissue between defect and

recipient site to accommodate flap base. High dissection of thoracodorsal donor vessels

III. Reduce recipient site complications. Use of proximal portion of muscle flap for calvarial and

cranioplasty coverage. Orientation of least reliable flap areas to minimize morbidity if

partial flap necrosis occurs. Vest-over-pants flap inset. Evenly distributed, low-pressure compression head dressing

IV. Reduce donor site seromas. Dependent, closed suction catheter drainage. Postoperative compression binder. Use of fibrin sealant

V. Improve aesthetic results. Tangential excision/debulking of muscle origin. Vest-over-pants inset to improve flap-scalp junction contour. Non-meshed skin grafts with quilting sutures


seromas resolved with percutaneous drainage and

continued use of a compression binder. There were

no other significant complications.

Except for the patient with radiation burns, all

patients had very good scalp contour. The

latissimus dorsi muscle atrophied to approxi-

mately 50% of its original thickness over the first

4 months, and the cranioplasty site did not show

significant contour deformity in any patient.

One patient had two synchronous areas of

local recurrence in the occipital scalp, posterior

to the latissimus dorsi scalp reconstruction,

7 months postoperatively. After radical resection

of the recurrences, the other latissimus dorsi

muscle was used for free flap reconstruction.

The mean follow-up period after reconstruc-

tion was 18 months (range, 8–27 months). Three

of the five patients died of their disease a mean of

13 months postoperatively (range, 8–18 months).

The remaining two patients were alive with no

evidence of disease at 26 and 27 months’ follow-up

(Table 1). The mean disease-free survival

for all five patients was 12.6 months (range,

3–27 months), and the mean overall survival was

18.2 months (range, 8–27 months).

DISCUSSION

Free tissue transfer is widely used for reconstruc-

tion of large scalp defects caused by oncologic

resection, because it enables placement of well-

vascularized tissue over the calvarium (fre-

quently devoid of periosteum) and the prosthetic

cranioplasty.2–10 However, despite the use of free

tissue transfer in scalp reconstruction, complica-

tions and poor aesthetic results often necessitate

surgical revision.

We had good outcomes and a low incidence of

complications with free latissimus dorsi flaps for

scalp reconstruction in our small series of mostly

elderly patients, all of whom had locally advanced

recurrent disease and received perioperative

radiotherapy. Several operative and perioperative

technical considerations likely reduced complica-

tions and improved flap survival and aesthetic

outcome (Table 2).

Techniques to Improve Outcome. Before insetting

the flap, the pedicle of the flap was carefully

passed through a wide tunnel to the recipient

vessel site. To prevent compression and kinking

of the pedicle within the tunnel, wide under-

mining between the scalp defect and the incision

site over the recipient vessels was carried out,

with scoring of the galea parallel to the pedicle

axis. Alternately, the tissue overlying the pedicle

was incised and undermined to accommodate the

proximal aspect of the muscle flap.

The pedicle length requirement was reduced,

and vein grafts were avoided by using the super-

ficial temporal vessels as recipients whenever the

orientation of the flap allowed (Figure 2). Despite

FIGURE 2. The superficial temporal vessels are used formicrovascular anastomoses when available and near the defect.

This often minimizes the pedicle length required and avoids the

need for vein grafts.

FIGURE 3. (A) Scalp and calvarial defect after resection in a 66-year-old man with recurrent squamous cell carcinoma of the occipital

scalp and calvarium. (B) The external carotid system and internal jugular vein were used as recipient vessels to minimize pedicle length. Ascalp incision and wedge excision were performed into which the flap base was inset, allowing the vascular pedicle of the flap to reach the

recipient vessels without vein grafts. Debulking of the muscle insertion area was performed to improve the contour. Considerable muscle

atrophy can be expected to occur over time. (C) Posterior and 1(D) lateral views of the scalp reconstruction 6 months postoperatively,

demonstrating a smooth scalp contour without excessive bulk at the flap base.


reports of the superficial temporal vein being

insufficient for microvascular anastomosis in a

number of cases, 4 we found that the superficial

temporal vessels, if palpable preoperatively, were

consistently reliable with adequate caliber in all

of the five cases in which they were used. In one

case, the neck vessels were used as recipients,

because the scalp defect extended posteriorly and

a shorter pedicle length was possible by coursing

the pedicle posterior to the ear. The thoracodorsal

vessels were dissected completely to their origin

from the axillary vessels if required, the pedicle

was dissected intramuscularly to increase pedicle

length when needed, and adjacent scalp tissue

between the defect and recipient vessels was

selectively incised and/or excised to accommodate

the muscle inset (Figure 3). The most proximal

aspect of the muscle flap near its tendinous

insertion was carefully resected or debulked

tangentially to facilitate insetting and improve

the resulting contour and aesthetic outcome.

Calvarial exposure after scalp reconstruction

in cancer patients is a complication that fre-

quently requires reoperation.4 To reduce the risk

of flap inset dehiscence leading to exposure of the

calvarium or the cranioplasty, as well as to

improve the contour of the flap-scalp junction,

we performed circumferential subgaleal wound

edge undermining for approximately 2 cm. A vest-

over-pants flap inset was used (Figure 4) by

suturing the muscle flap edge to the galea of the

undermined scalp wound edge to provide an

approximately 1.5-cm zone of ‘‘scalp-over-muscle’’

coverage around the entire circumference of the

inset flap. This maneuver improved the aesthetic

outcome by preventing the depression often seen

at the scalp-muscle junction and reduced the risk

of wound dehiscence in the event of partial

muscle edge necrosis. The dermis of the scalp

edge was then secured to the surface of the

latissimus dorsi muscle flap with interrupted

absorbable sutures before the placement of a

nonmeshed, split-thickness skin graft. When

possible, the most proximal aspect of the muscle

was used for calvarial and cranioplasty coverage

and the most distal (least reliable) aspect of the

flap was positioned so that the least morbidity

would result if partial flap loss were to occur. For

example, the distal aspect of the flap was

commonly placed over intact temporalis muscle

and/or intact periosteum, remote from the pros-

thetic calvarial bone reconstruction.

A closed-suction drain was placed under the

muscle flap, away from the cranioplasty site or

vascular pedicle and oriented parallel to the flap’s

vascular axis. This parallel orientation reduces

the risk of vascular compromise to flap areas

distal to the drain location (directly over bone),

particularly if a compressive dressing is used to

help immobilize the skin graft.

FIGURE 4. A vest-over-pants inset is used to prevent subsequent depression and/or bone exposure at the flap-scalp junction. The

periphery of the scalp defect is elevated through the subgaleal plane for approximately 2 cm, the muscle edge is sutured to the galea

beneath the undermined scalp, and then the scalp defect edge is secured over the muscle flap surface.


To optimize skin graft survival, we used non-

meshed skin grafts with multiple 6-0 chromic

quilting sutures (Figure 5), meticulous skin graft

insetting with interrupted and running chromic

sutures, and carefully constructed, uniformly

compressive, low-pressure head dressings. The

dressing we found most useful was N-TerFace

(Winfield Laboratories, Dallas, TX), covered by

sterile cotton and an elastic ‘‘fishnet’’ tubing to

secure the dressing. Transcutaneous Doppler

arterial and venous signals were marked with a

Prolene suture and exposed by separating the

cotton and cutting a hole in the N-TerFace for the

ultrasound transducer. The skin-grafted flap could

also be directly visualized through this opening.

One of the most important postoperative

considerations is proper patient positioning to

avoid compression of the flap or pedicle, which are

both located directly over bone or bony recon-

struction. Our patients were positioned with the

head of the bed elevated 45 to 70 degrees and the

patients’ heads maintained in a neutral position

using pillows or foam pads placed on both sides of

and behind the head, inferior to the occiput. This

was sufficient in most cases to prevent compres-

sion on the flap site, even during sleep. If this is

ineffective in preventing flap compression, partic-

ularly with posterior reconstructions, a halo

device should be considered.

To reduce the risk of seromas, two closed-

suction drainage catheters were positioned in the

donor site to provide dependent drainagewhile the

patient was supine or upright. A circumferential

compression binder was worn postoperatively over

the donor site area until all drainage catheters had

been removed for at least 3 weeks. Each drain was

removed when its 24-hour output was less than

25 mL per drain for 2 consecutive days.

Recently, we have started using aerosolized

fibrin sealant (Tisseel, Baxter Healthcare Corp.,

Deerfield, IL) sprayed into the latissimus dorsi

muscle donor defect immediately before closure;

in our preliminary experience, we have observed

less drainage fluid and a lower incidence of

seromas with this technique. The thrombin

component of a 5-mL Tisseel kit is diluted 100-

fold (from 500 to 5 IU/mL per manufacturer’s

dilution instructions). This dilution allows addi-

tional time for approximating the edges of the

donor site wound before polymerization. After

application of the fibrin sealant, several sutures

preplaced through the wound edges are immedi-

ately tied to allow the fibrin sealant to set with

the donor site tissues in the proper position.

CONCLUSIONS

The free latissimus dorsi muscle flap with a split-

thickness skin graft is an excellent, reliable

option for reconstruction of large scalp defects.

We believe that certain technical aspects contrib-

ute to a good functional and aesthetic outcome,

even in elderly patients with advanced disease

receiving perioperative radiotherapy.

REFERENCES

1. Robson MC, Zachary LS, Schmidt DR, Faibisoff B,Hekmatpanah J. Reconstruction of large cranial defectsin the presence of heavy radiation damage and infectionutilizing tissue transferred by microvascular anastomoses.Plast Reconstr Surg 1989;83:438–442.

2. Furnas H, Lineaweaver WC, Alpert BS, Buncke HJ. Scalpreconstruction by microvascular free tissue transfer. AnnPlast Surg 1990;24:431–444.

FIGURE 5. Nonmeshed split-thickness skin grafts are meticu-

lously inset over the muscle flap with edge eversion and secured

to the flap with absorbable quilting sutures.


3. Lutz BS, Wei FC, Chen HC, Lin CH, Wei CY. Reconstruc-tion of scalp defects with free flaps in 30 cases. Br J PlastSurg 1998;51:186–190.

4. Hussussian CJ, Reece GP. Microsurgical scalp reconstruc-tion in the patient with cancer. Plast Reconstr Surg2002;109:1828–1834.

5. Earley MJ, Green MF, Milling MA. A critical appraisal ofthe use of free flaps in primary reconstruction of combinedscalp and calvarial cancer defects. Br J Plast Surg 1990;43:283–289.

6. McLean DH, Buncke HJ, Jr. Autotransplant of omentumto a large scalp defect with microsurgical revasculariza-tion. Plast Reconstr Surg 1972;49:268–274.

7. Pennington DG, Stern HS, Lee KK. Free-flap reconstruc-tion of large defects of the scalp and calvarium. PlastReconstr Surg 1989;83:655–661.

8. Ueda K, Harashina T, Inoue T, Tanaka I, Harada T.Microsurgical scalp and skull reconstruction using aserratus anterior myo-osseous flap. Ann Plast Surg 1993;31:10–14.

9. Lee B, Bickel K, Levin S. Microsurgical reconstruction ofextensive scalp defects. J Reconstr Microsurg 1999;15:255–262.

10. Lutz BS. Aesthetic and functional advantages of theanterolateral thigh flap in reconstruction of tumor-relatedscalp defects. Microsurgery 2002;22:258–264.


enhancing the outcome of free latissimus dorsi muscle flap

Documents