CHAPTER 60 LATISSIMUS DORSI FLAP BREAST RECONSTRUCTION
DENNIS C. HAMMOND AND MICHAEL A. LOFFREDO
Reconstruction with autologous tissue provides the patient with a reconstructed breast created with her own tissues, obviating the potential complications associated with a prosthesis. The disadvantage with this strategy is related to the creation of an additional donor site with scarring and potential morbidity. Although reconstruction with tissue expanders and implants eliminates the need for the additional donor site, the potential complications associated with these devices are a concern. The latissimus dorsi musculocutaneous flap (LDF) seems to offer no advantage as, most commonly, tissue expanders and implants are still required, and the additional donor site is created on the back. For this reason, the LDF remains a distant third option for many reconstructive breast surgeons. With the development of newer and more effective tissue expanders and implants, however, the advantages of combining these devices with the well-vascularized LDF have generated a resurgence of interest in the technique. This chapter focuses on the technical strategies for optimizing the use of the LDF with tissue expanders and implants.
The LDF was originally described in 1906 by Iginio Tansini in Italy.1 It was used to reconstruct mastectomy wounds at the time, but soon fell from favor, to be rediscovered in the late 1970s.2,3 Since its rediscovery, the flap has been used to reconstruct nearly every part of the body, as both a pedicled and a free flap. The LDF is a reliable and richly vascularized flap, and the proximity of the flap to the anterior chest wall makes it an ideal choice for providing the muscle, fat, and skin for use in reconstructing the breast after mastectomy. Sacrifice of the muscle creates a negligible functional deficit except in extremely athletic women.4-10 Using this as a basic strategy, there are several variables both in flap design and elevation, as well as in expander and implant choice, which can be manipulated to maximize the aesthetic quality of the result, while minimizing the donor-site morbidity and potential complications.
Historically, use of the LDF has involved transposing only the muscle to the mastectomy defect with an isolated island of skin and fat of varying size positioned on top of the muscle. Although this can effectively provide cutaneous cover for the breast, a more effective technique for providing volume is to harvest the deep layer of subcutaneous fat with the muscle beyond the skin island. The deep thoracic fascia provides a readily recognizable anatomic landmark that guides dissection and even allows the deep fatty layer below the fascia to be harvested beyond the borders of the muscle. By increasing the overall volume of the flap and creating a volume-added latissimus flap, the ability to fill in and soften the margins of the mastectomy defect is enhanced.
Recent developments in expander and implant design have resulted in a wide array of devices available for use in reconstruction. Choosing between round and anatomically shaped devices that are either textured or smooth and filled with either saline or silicone provides a variety of choices, which can be strategically exploited to solve individual reconstructive problems. For instance, a thin patient with stark breast contours may be served best by an anatomically shaped silicone gel textured implant. By combining the volume-added latissimus flap with an appropriately chosen expander or implant, excellent results are possible.
With the patient in the standing position, the borders of the latissimus dorsi muscle are delineated. The midline of the back is identified and the tip of the scapula is marked (Figure 60.1A). With the arm elevated, the anterior border of the muscle is drawn as it extends from the posterior border of the axilla downward toward the iliac crest. The upper border of the muscle is drawn as it extends from the axilla over the tip of the scapula to the midline of the back (Figure 60.1B). The origin of the muscle is marked inferiorly as it curves from the lower midline of the back to the anterior border of the muscle. The inferior segment of the trapezius is drawn as it overlaps the upper medial border of the latissimus as a reminder of this important anatomic relationship.
Once the limits of the muscle have been identified, the location and orientation of the skin island is identified. The skin island is positioned in the center of the muscle to ensure equal soft-tissue coverage of the expander or implant in all directions once the skin island is inset. When a small, circular skin island is required, as is commonly the case when a skin-sparing mastectomy strategy is used, the skin island is positioned directly in the center of the flap. The relaxed skin tension line is identified as it passes through the center of the skin island, and this line guides the drawing of a gentle ellipse around the skin island (Figure 60.1C). By tapering off around the skin island medially and laterally, adequate exposure for flap dissection is provided, while allowing for direct closure of the skin defect without dog-ear formation at the medial or lateral ends of the incision. In this patient, the skin island also includes the addition of an immediate nipple–areola reconstruction using a skate flap purse-string technique (Figure 60.1D).11 Alternatively, the skin island can be left intact to reconstruct the nipple–areola at the second stage.
When a larger skin island is required, as in cases of delayed breast reconstruction, the same strategy is employed, attempting to place the elliptical skin island more or less in the center of the flap and orienting the long axis of the ellipse along the same relaxed skin tension line. The advantage of orienting the long axis of the skin paddle in this fashion is that, despite the scar sweeping up to the upper back, it heals in an acceptable fashion. This is in contradistinction to other scar orientations, which can be quite unsightly as the orientation of the skin island crosses the relaxed skin tension lines of the back, resulting in widened or hypertrophic scars.
FIGURE 60.1. Preoperative markings. A. The midline of the back is marked along with the tip of the scapula. B. With the arm raised, the sweep of the superior border of the latissimus muscle can be drawn in as it courses over the tip of the scapula to the midline of the back. The anterior border of the muscle is identified and marked as it runs inferiorly from the posterior border of the axilla to the iliac crest. C. The center of the muscle is identified and the relaxed skin tension line, which passes through this point, is drawn. This line generally sweeps from superomedial and curves anteriorly across the back toward the abdomen. Placing the incision for the skin island in this line results in the least visible postoperative scar. D. A gentle ellipse is drawn around the circular skin island, tapering off medially and laterally so as to provide a smooth postoperative scar. E. On the lateral view, the zone of adherence marked by the X’s should be respected and these tissues should not be elevated during flap transfer. Instead, the flap is optimally passed through a tunnel created high in the axilla and dropped into the mastectomy defect. This preserves the lateral breast contour, which is a landmark that can be difficult to create with internal sutures.
Once the skin island is marked and the limits of the muscle identified, the zone of attachment between the upper anterior border of the muscle and the lateral border of the breast is marked. If at all possible, avoid dissection in this area as this preserves the lateral border of the breast. This contour can be difficult to reconstruct with suture plication if it is released during either the mastectomy or in making the tunnel for passage of the LDF. Finally, the tunnel through which the flap is to be passed anteriorly is drawn high in the axilla. This serves to preserve the lateral contour of the breast and limits any potential lateral chest wall fullness or deformity that can occur as a result of overzealous tissue release in this area (Figure 60.1E).
After the mastectomy is completed, the viability of the skin flaps is assessed. If there is a question of mastectomy flap ischemia, fluorescein is infused to assess vascular perfusion. Debridement of all nonperfused areas is performed. It is helpful at this point to predissect the high axillary tunnel through the mastectomy defect in preparation for subsequent flap transfer. This space is then easily entered during flap dissection on the back, facilitating subsequent transposition of the flap once it is completely released, preventing inadvertent overdissection of the lateral chest wall during creation of the tunnel.
At this point, the wound is temporarily stapled closed and covered with a sterile dressing. The patient is rotated into the prone position for both unilateral and bilateral reconstructions. In the prone position, the anatomic landmarks are more easily identified than in the lateral decubitus position.
The skin island is incised and the thoracic fascia is divided along the line of incision. With division of the fascia, the wound springs open as the loose, deep, fatty layer is exposed. Dissection then proceeds just under the fascia in all directions, keeping the deep layer of fat attached to the muscle. Once the limits of the muscle are reached, dissection proceeds through the deep layer of fat down to the margins of the latissimus muscle in all directions. The trapezius muscle is identified in the upper medial corner of the dissection space and the latissimus is released from under it. The upper border of the latissimus muscle is then identified and released. The fibers of origin are then peeled away from the fascia of the back, extending from the upper medial corner of the dissection space toward the iliac crest. As the muscle is undermined, crossing perforators from the intercostal spaces are controlled. Inferiorly, the muscle is divided from its attachments to the iliac crest. The muscle is released from its attachments to the serratus anterior and the anterior border is identified and released, care being taken to avoid inadvertent elevation of the fibers of origin of the external oblique in the anterior-inferior corner of the dissection. The muscle is then elevated toward the axilla and any remaining attachments to the teres major muscle are divided.
Communication with the mastectomy wound is then made high in the axilla and the muscle is passed anteriorly to the mastectomy defect. The pedicle is easily identified entering the underside of the muscle and care is taken to avoid tethering or injuring the vascular leash. With full release of the muscular attachments, there is no need to divide the serratus branch, which can become important, as this vascular conduit can support the flap entirely if the main thoracodorsal pedicle was injured during the original mastectomy.12
The back wound is then closed over suction drainage. The placement of quilting sutures securing the upper and lower back flaps to the intercostal muscles may assist in preventing or limiting persistent postoperative drainage from the back donor site.13 The patient is rotated back into the supine position. The flap is fully withdrawn into the mastectomy wound and the insertion of the muscle identified. I prefer to divide this insertion completely, just above the entry point of the vascular pedicle, as this facilitates easy rotation of the LDF in any direction required, depending on the dimensions of the mastectomy defect. Alternatively, it is a reasonable compromise to divide the posterior 90% of the insertion, as the remaining attachments protect against inadvertent traction being placed on the pedicle, and yet full transfer of the flap into the mastectomy defect is greatly facilitated.
The vascular pedicle is readily identified. I prefer to divide the thoracodorsal nerve at this point. With division of the nerve, unwanted and distracting motion in the reconstructed breast is greatly diminished or eliminated, a finding that has been a welcome addition to my results over the years. Such denervation has not resulted in enough volume loss in the flap to significantly detract from the aesthetic result of the reconstruction.
The flap is now prepared for insetting. In small- to medium-sized breasts, there is no need to elevate the pectoralis major muscle. The edges of the LDF are simply sutured into the medial, superior, and lateral margins of the mastectomy defect, and the tissue expander is placed under the flap. The remaining edge of the LDF is then inset into the inferior margin of the mastectomy wound around the inferior border of the expander, providing complete muscle coverage for the device.
In larger breasts, the surface area provided by the LDF may not be enough to provide sufficient padding for the entire surface area of the mastectomy defect without tethering, and the skin envelope of the breast may not be sufficiently filled out. In these cases, full release of the pectoralis major muscle can enlarge the muscular soft-tissue envelope, allowing the now upwardly retracted pectoralis muscle to cover the upper portion of the defect, and using the LDF to cover the lower portion. Centrally, where these two muscles meet, the edges are simply sutured together in a vest-over-pants fashion, with the latissimus secured on top of the pectoralis major. Whatever strategy is used, it is important to close off the tunnel leading to the back donor site to an opening of only 2 to 3 cm to prevent inadvertent slippage of the expander through the axilla into the back postoperatively.
At this point, the dimensions of the skin island are finalized and inset, discarding redundant skin as needed. In cases of immediate reconstruction, often only a circular skin island is required to fit into the defect created by removal of the nipple and areola. In cases of immediate nipple–areola reconstruction, the skate flap purse-string technique is utilized prior to suturing the skin island into place (Figure 60.2). In delayed reconstruction, an ellipse is generally used to fill in the cutaneous defect created at the time of mastectomy. This ellipse of skin can be positioned in one of two ways. Perhaps the most straightforward inset strategy involves opening the mastectomy wound along the scar line and insetting the flap directly into the resulting wound. In this manner, no new scars are created on the breast, and the vascularity of the mastectomy flaps is not compromised. In thin patients, or in patients with mastectomy scars positioned low on the chest, it is possible to resect nearly the entire lower mastectomy flap and replace it with the LDF skin island, placing the lower scar directly in the inframammary fold. However, in patients with larger breasts or high-riding mastectomy scars, not all of the lower mastectomy flap can be removed, and the LDF skin island will create an obvious patch effect once it is inset into the reconstructed breast. To prevent this, many surgeons will ignore the previous mastectomy scar and open the mastectomy wound by making an incision low and lateral along the proposed inframammary fold. Once the LDF skin island is then inset, the lower scar will be hidden in the fold, and the shape of the ellipse will assist in creating a rounded ptotic appearance to the reconstructed breast. This approach can risk compromise of the vascularity of the remaining lower mastectomy flap as a result of the crisscrossing scars; for this reason, some surgeons will keep the remaining upper mastectomy flap attached to the pectoralis major muscle, positioning the expander under the muscle. The disadvantage is that the breast may have unwanted motion postoperatively because of the subpectoral placement.
After positioning the expander and insetting the flap and the skin island, the skin incisions are closed over suction drains. It should be noted that in cases where adjustment of the opposite breast is planned, that procedure (whether it is breast augmentation, reduction, or mastopexy) is performed at this initial stage. In this fashion, the breast is allowed to settle until the stage 2 procedure is performed, which enhances the accuracy of the second procedure, as the reconstructed breast is matched to a stable opposite breast size and shape.
Postoperative recovery is generally uneventful, with most patients leaving the hospital within 2 to 3 days. Early motion of the arms and shoulders is encouraged to prevent stiffness. Expander inflation is performed in the office setting as needed to achieve the desired final volume, beginning as early as 2 weeks postoperatively. Often only one or two expansions are necessary because of the adequacy of the dimensions of the skin surface area created by adding the latissimus skin island. For this reason, it is also not necessary to overinflate the expander to a significant degree in most cases.
After the recovery from the first stage is complete and all swelling has resolved, the final shaping of the breast is performed, usually 4 to 6 months later. The procedure is generally performed in the outpatient setting. At this stage, the tissue expander is removed, the breast is reshaped as needed, and the nipple and areola are reconstructed if not performed during the first stage. In cases of immediate nipple– areola reconstruction, this second-stage procedure can serve as an opportunity to make any minor revisions as necessary to improve upon the previously reconstructed nipple and areola. Additional adjustments to the opposite breast can also be made as needed. Under the best of circumstances, all that will be required to complete the reconstruction will be to remove the tissue expander, replacing it with a permanent implant, and reconstruction of the nipple and areola using one of several different techniques. It bears noting that, because the back dermis is quite thick, the appearance and longevity of nipples made with latissimus flap skin tend to be excellent. In selected cases, further modification of the reconstructed breast may be required to obtain the optimal result. These modifications may include contour reconfiguration with elevation or lowering of the inframammary fold, widening of the pocket, or capsulectomy with removal of scar. Breast implant dimensions, volume, and shape are chosen to give the best possible result. Liposuction of the lateral chest wall is occasionally required to treat excess fullness in this area. Using this staged approach, excellent results can be obtained in most cases.
FIGURE 60.2. Flap inset and immediate nipple–areola reconstruction. A. The flap has been passed anteriorly through the axilla and the tissue expander positioned centrally within the mastectomy defect. B. The muscle is then wrapped around the expander, suturing the edges of the muscle into the margins of the mastectomy defect. This maneuver softens the contours of the mastectomy wound, improving the overall quality of the reconstructive result. C. Full-thickness incisions are made around the areolar hemiflaps and the skate flap. The skate flap is then elevated and assembled keeping a uniform thickness of fat on the underside of the flap. D. The two areolar hemiflaps are then sutured together and the entire nipple–areola construct is then sutured into the defect using a periareolar purse-string suture. E. After inset of the skin island in this case of a bilateral reconstruction performed in conjunction with a periareolar skin-sparing mastectomy, an aesthetic result has been created, even at this early stage, with the tissue expanders in place.
EXPANDER VERSUS IMPLANT
Many surgeons insert the primary implant at the same time the latissimus flap is inset. This is an attractive option, as the need for a second separate procedure to remove the tissue expander is avoided. It can be accomplished, particularly in cases of immediate breast reconstruction, because generally there is no loss of skin envelope surface area to the breast. Consequently, there is no need to “expand” the skin. Despite this, I prefer to use a tissue expander during the stage 1 procedure. When true expansion is not required, I consider the expander to be an intelligently chosen spacer, holding the soft-tissue envelope of the reconstructed breast open while the soft tissues heal around it. This strategy facilitates more accurate implant selection at the stage 2 procedure, allowing changes in implant base diameter, projection, volume, and shape to be made based on how the breast has recovered from the stage 1 procedure. As well, a strategy for manipulation of the soft-tissue pocket at stage 2 is built into the operative plan, which only enhances the ultimate result.
The LDF is an excellent option for almost any reconstructive situation. It can be used with ease in cases of immediate or delayed reconstruction (Figure 60.3). It is a particularly attractive option for patients who require bilateral reconstruction (Figure 60.4). In patients who may not have enough volume for a bilateral transverse rectus abdominis musculocutaneous (TRAM) flap and who want more of an expander/implant reconstruction, the bilateral latissimus flap option is an excellent choice. The LDF is also a useful flap for reconstruction of the partial mastectomy defect or for autogenous salvage in cases of significant fat necrosis after TRAM flap reconstruction. At times, in patients with relatively small opposite breasts, it can be used as a completely autogenous flap, obviating any need for an implant. In these latter circumstances, the “volume-added” strategy becomes a particularly attractive option for increasing the available soft-tissue bulk provided by the flap.
The most recognized complication of latissimus flap harvest is donor-site seroma formation. It is not uncommon for the drain to remain in the back donor site for as long as 6 weeks postoperatively. Occasionally, outpatient aspiration of a persistent seroma is required after drain removal. With time, fluid accumulation eventually stops, leaving behind an empty bursa of varying dimensions. Although isolated arm strength can be diminished after flap transfer, this is rarely a significant finding, as the other muscles of the back compensate for the absent muscle. The vascularity of the LDF itself is robust, and only rarely is there any vascular compromise or fat necrosis noted after LDF transfer. However, marginal skin necrosis at the donor site can occur particularly in smokers, which might suggest avoiding excessive undermining. After transfer of the flap with placement of a tissue expander or implant, capsular contracture can occur. Treatment generally consists of complete capsulectomy, which alleviates the problem in the vast majority of cases.
FIGURE 60.3. Unilateral reconstruction. A, B. Preoperative appearance of a 54-year-old woman scheduled to undergo a left-sided modified radical mastectomy for adenocarcinoma of the breast. C, D. One-year postoperative appearance after eventual placement of a 400-cc smooth, round, silicone gel implant on the left, and an augmentation of the right breast with a 375-cc smooth, round, silicone gel implant. The reconstructed nipple and areola have been tattooed.
FIGURE 60.4. A. Preoperative appearance of a 59-year-old woman with lobular carcinoma in situ of the left breast. She had undergone subcutaneous mastectomy with implant reconstruction in the remote past. B. The preoperative marks outline a periareolar skin-sparing mastectomy that includes a lateral extension for better access to the breast. C. The back marks outline the circular skin island used to replace the missing nipple and areola. The position of the ellipse is oriented along the relaxed skin tension lines. D. Appearance 6 months after her stage 1 procedure reveals good symmetry and a pleasing overall aesthetic result. E. The preoperative marks for her stage 2 procedure outline her bilateral nipple and areola complex reconstruction along with soft-tissue recontouring in the upper outer quadrant of the breast. The tissue expanders will be replaced with silicone gel implants at a second stage. F. Appearance 1 year after placement of a 600-cc smooth, round, silicone gel implant on the right, and a 500-cc implant on the left. G. Appearance of the back reveals no contour deformity and a well-healed scar with no distortion or bunching.
The LDF provides a readily available and reliable block of tissue that can be used in breast reconstruction. Because it dramatically enhances the ability of the plastic surgeon to artistically reconstruct the contours of the breast in a wide variety of clinical settings, it is recommended as an essential technique in the armamentarium of the reconstructive breast surgeon.
1. Maxwell GP. Iginio Tansini and the origin of the latissimus dorsi musculocutaneous flap. Plast Reconstr Surg. 1980;65:686.
2. Olivari N. The latissimus flap. Br J Plast Surg. 1976;29:126.
3. Schneider WJ, Hill HL, Brown RG. Latissimus dorsi myocutaneous flap for breast reconstruction. Br J Plast Surg. 1977;30:277.
4. Laitung JKG, Peck F. Shoulder function following the loss of the latissimus dorsi muscle. Br J Plast Surg. 1985;38:375.
5. Russell RC, Pribaz J, Zook EG, et al. Functional evaluation of latissimus dorsi donor site. Plast Reconstr Surg. 1986;78:336.
6. Fraulin FOG, Louie G, Zorrilla L, et al. Functional evaluation of the shoulder following latissimus dorsi muscle transfer. Ann Plast Surg. 1995;35:349.
7. Luce PA, Hammond DC. Latissimus dorsi musculocutaneous flaps and tissue expanders/implants in immediate breast reconstruction. Plast Surg Forum. 1995;64:133.
8. Fisher J, Hammond DC. The combination of expanders with autogenous tissue in breast reconstruction. Clin Plast Surg. 1994;21:309.
9. Hammond DC. Postmastectomy reconstruction of the breast using the latissimus dorsi musculocutaneous flap. Cancer J. 2008;14:248.
10. Hammond DC. Latissimus dorsi flap breast reconstruction. Plast Reconstr Surg. 2009;124:1055.
11. Hammond DC, Khuthaila D, Kim J. The skate flap purse-string technique for nipple-areola complex reconstruction. Plast Reconstr Surg. 2007;120:399.
12. Fisher J, Bostwick J III, Powell RW. Latissimus dorsi blood supply after thoracodorsal vessel division: the serratus collateral. Plast Reconstr Surg. 1983;72:502.
13. Rios J, Adams WP, Pollock T. Progressive tension sutures to decrease latissimus donor site seroma. Plast Reconstr Surg. 2003;112:1779.