BREAST
Latissimus Dorsi Flap Salvage of the Partially Failed TRAM Flap Breast Reconstruction
Dennis C. Hammond, M.D.
Background: The transverse rectus abdominis musculocutaneous (TRAM) flap is Amy M. Simon, M.D.
an excellent option for patients desiring autogenous breast reconstruction. How-Dana K. Khuthaila, M.D.
ever, partial flap ischemia can result in isolated areas of skin and fat necrosis, which, Liberty Hoberman, M.D.
when removed, can create irregular and sharply defined defects that distort the Sam Sohn, M.D.
aesthetic appearance of the reconstructed breast. Such defects can be difficult to
reconstruct with local-tissue rearrangement or with breast implants. Methods: A subset of patients undergoing TRAM flap breast reconstruction who experienced significant partial flap necrosis was identified. After debridement, 14 consecutive patients subsequently underwent revision reconstruction with an autogenous, volume-added latissimus dorsi musculocutaneous flap. Demographic data and results in this subgroup of patients were collected. Results: All patients in the study experienced significant deformity in the reconstructed breast as a result of the removal of the necrotic tissue. The irregular and sharply defined defect was filled in completely and smoothly with the autogenous latissimus dorsi flap, obviating the need for an implant. The flap survival rate was 100 percent as the latissimus dorsi flap provided a reliable volume of soft tissue with which to fill in the defect in the breast, thus preserving the autologous nature of the reconstruction. Conclusion: Using the latissimus dorsi musculocutaneous flap as soft-tissue filler enables restoration of appropriate volume and an aesthetic shape to the reconstructed breast in patients who have developed fat necrosis after TRAM flap breast reconstruction. (Plast. Reconstr. Surg. 120: 382, 2007.)
Grand Rapids, Mich.
he transverse rectus abdominis musculocutaneous flap (TRAM) flap is an excellent option for select patients seeking breast reconstruction with autogenous tissue.1,2 Advantages include artistic flexibility in breast shaping, a natural consistency and feel to the reconstructed breast, and avoidance of any of the potential complications associated with the use of breast implants. An autogenous flap can also provide well-vascularized skin, muscle, and fat, which is especially advantageous when reconstructing a patient who has had postoperative irradiation after complete or partial mastectomy.3,4
One common complication associated with TRAM flap breast reconstruction is partial flap necrosis.5 If large areas of skin or fat fail to survive, the aesthetic result can be compromised
From the Center for Breast and Body Contouring; Michigan State University; and private practice. Received for publication December 19, 2005; accepted March 21, 2006. Copyright ©2007 by the American Society of Plastic Surgeons
DOI: 10.1097/01.prs.0000267327.88506.54
once these necrotic areas are removed. Subsequent salvage reconstruction with local tissue rearrangement or breast implants can be difficult because the resulting defect is often irregular and can have sharply defined borders.
In these cases, the use of an autogenous latissimus dorsi musculocutaneous flap is an attractive option for salvage reconstruction.6,7 The flap is a proven reconstructive option, has a reliable vascular supply, provides an adequate volume of richly vascularized tissue to reconstruct even large defects, and has sufficient pliability to be molded to fit smoothly into even the most irregular of partial mastectomy defects. We present here the results of 14 patients who had significant, deforming partial flap necrosis after immediate TRAM flap breast reconstruction who underwent salvage reconstruction using an autogenous latissimus dorsi flap.
PATIENTS AND METHODS
Records were reviewed retrospectively to identify patients who had undergone a pedicled TRAM flap breast reconstruction following mastectomy during the 10-year period from 1992 to 2002.
These were then narrowed to include only those patients who had developed necrosis of the skin or subcutaneous tissue of the TRAM flap, followed by excision and additional salvage reconstruction with a latissimus dorsi flap. Fourteen consecutive patients were included in the study. Charts were reviewed to identify patient age, medical history, comorbidities, and radiation history. The percent of TRAM flap lost, location of the defect, and latissimus dorsi flap donor-site complications were noted. Patient satisfaction with the procedure was also evaluated.
Technique
Initial Management
Partial TRAM flap necrosis often involves both skin and fat. In cases in which a skin-sparing mastectomy strategy was used, the portion of the flap that ultimately becomes necrotic can be sequestered underneath an otherwise viable mastectomy flap, where it becomes known only after it forms into a palpable mass. When this happens, a surrounding zone of localized edema also develops. It can take many months for this swelling to resolve and for the area of necrotic fat to become encapsulated by scar. For this reason, it is advisable to wait until the extent of fat necrosis is clearly defined and stable before attempting excision with reconstruction. Removing such areas before the area of fat necrosis has matured is difficult and can result in unnecessary removal of surrounding viable tissue with the creation of a more extensive defect than would otherwise be required. This maturation process can take between 6 months and 1 year. In cases in which a larger skin paddle from the TRAM flap was used to assist in reconstructing the skin envelope of the reconstructed breast, partial skin necrosis is a more pressing concern because the necrotic portion of the flap may be exposed. Here, the extent of visible skin necrosis is allowed to mature over 7 to 14 days or until the borders of the necrotic portion of the flap become well defined, at which time debridement is performed by removing the skin eschar along with any obvious areas of underlying necrotic fat. This is done to limit the amount of time the wound remains open. It must be noted, however, that, because of tissue swelling, the differentiation between necrotic fat and potentially viable fat can be difficult to make because of residual edema in the flap, particularly in the zone of ischemia. For this reason, it is advisable to be conservative with tissue removal, limiting debridement to only those areas that are obviously without blood supply and then closing the wound. Complete debridement of any residual necrotic areas can be performed once the wound has stabilized.
Once the area of necrotic fat is clearly identified and/or the wound is stable in size, the effect of any subsequent debridement on the shape of the breast must be assessed. Small areas of fat necrosis can often be debrided with little to no effect on breast shape. However, larger areas can create irregular and isolated defects in the breast with sharp edges that are not amenable to simple tissue rearrangement or placement of breast implants. In these cases, the latissimus dorsi musculocutaneous flap is used to fill in the resulting defect and restore symmetric and aesthetic contours to the reconstructed breast.
Latissimus Dorsi Musculocutaneous Flap Elevation
The volume of the missing segment of the breast is estimated and an appropriate segment of latissimus dorsi musculocutaneous flap is identified so as to completely fill in the defect. Volume deficits are managed easily without restriction with regard to flap rotation; however, the need for skin must be assessed carefully so that the skin island on the latissimus dorsi flap can be positioned properly. To maximize the aesthetic quality of the back scar, the skin island is drawn along the axis of the lines of skin tension as they curve anteriorly across the torso. For central to lateral breast defects, the skin island is located in the center of the flap. For medial defects, the skin island is moved inferiorly 2 cm toward the origin of the latissimus to allow it to reach easily without tension once the flap is transferred to the chest wall. In cases in which no skin island is required, a single-access incision across the midportion of the flap facilitates flap elevation without the need for including a segment of skin.
In an effort to increase the volume of the flap, the deep layer of fat on the back can be harvested with the flap during flap elevation. This fatty layer is identified easily deep to the thoracic fascia, which springs open once this fascia is surgically divided. The deep layer of fat is then separated easily from the underside of the fascia, and it is kept attached to the muscle during flap elevation. It is possible to harvest this deep fat as an extension off the peripheral margin of the muscle as a means to increase the volume of the flap. As well, the subserratal fat pad is harvested with the flap to add more volume to the latissimus dorsi flap.8,9
To allow freedom in flap rotation on the chest wall, the insertion of the muscle into the humerus is divided for 90 percent of the posterior attach
Plastic and Reconstructive Surgery • August 2007
| Percentage of | Quadrant of TRAM | Quadrants | |||
|---|---|---|---|---|---|
| Patient | Age (yr) | Type of TRAM | TRAM Lost | Necrosis in Breast | Involved |
| 1 2 3 4 5 6 7 8 9 10 11 12 13 14 | 52 33 48 58 54 55 41 66 49 56 42 45 40 37 | Unipedicle-bilateral Unipedicle Bipedicle Unipedicle Unipedicle-bilateral Unipedicle Bipedicle Unipedicle Unipedicle Unipedicle-bilateral Unipedicle Unipedicle Unipedicle Unipedicle | 75 50 75 25 25 25 25 25 25 25 25 50 50 25 | S, M, I, L IM, IL C, IL, IM IL, IM SL SL IL IM IM IL IL, IM SL, IL IL, IM IL, IM | 4 2 3 2 1 1 1 1 1 1 2 2 2 2 |
ment. This allows the flap to advance 10 to 12 cm out of the axilla and facilitates insetting of the skin island in virtually any orientation as needed. Keeping the last 10 percent of the insertion intact prevents inadvertent traction from being placed on the vascular pedicle and avoids any potential injury to the vessels.9 At this point, the thoracodorsal nerve can be identified and, if desired, divided to prevent unwanted and distracting animation of the reconstructed breast. Toward the end of this study period, sectioning the thoracodorsal nerve became a standard maneuver in our latissimus flap elevation technique. Once the flap is completely elevated, it is tucked up into the axilla and the back donor site is closed over suction drains.
Latissimus Dorsi Flap Insetting
If not already accomplished, the area of necrotic skin and fat in the TRAM flap are debrided thoroughly until only soft and well-vascularized tissue remains. Communication with the back is then made under the TRAM flap through a tunnel located high in the axilla. A concerted attempt is made to preserve the perforators running from the rectus abdominis muscle to the overlying skin. For medial defects, a tunnel superior to the rectus muscle and under the reconstructed breast is made, and the latissimus dorsi flap is pulled into the defect. For lateral defects, the tunnel is made lateral and inferior to the rectus muscle. Although revascularization from the surrounding tissue likely will support the remaining TRAM flap tissue, every effort is made to preserve the superior epigastric vessels as they pass over the ribs to the remaining TRAM flap. Once the latissimus dorsi flap is in place, it is folded and secured loosely into the defect as needed to completely fill in the void caused by removal of fat. In cases of skin deficit, the skin island is also inset to fully restore the skin
Table 1. Patient Summary
surface area of the reconstructed breast. In our experience, we have found it advisable to overcorrect the required volume of the reconstructed breast by an estimated 10 to 20 percent to allow for postoperative volume loss, which can affect the eventual size and shape of the breast. It is important to assess the shape of the breast with the patient in the upright position to reconstruct the aesthetic contours as accurately as possible. If excessive flap volume is present, the redundant residual flap is removed by excising fat along the distal peripheral edges. In this fashion, the axial blood supply will not be interrupted inadvertently. A drain is placed and the wound closed to complete the procedure.
RESULTS
A total of 14 patients underwent secondary breast reconstruction with a latissimus dorsi musculocutaneous flap following a partially failed TRAM flap (Table 1). The mean age was 48 (range, 33 to 66). Preoperatively, 2 patients were smokers, 2 were obese, 3 were hypertensive, and 5 had a history of preoperative radiation therapy. In 12 patients (86 percent), a unipedicle TRAM flap was used; in 2 patients (14 percent), a bipedicle flap was used. Also, 3 patients (21 percent) underwent bilateral breast reconstruction with a unipedicle TRAM to each side. The volume of TRAM flap loss was estimated at 25 percent or less in 9 patients (64 percent), 25 to 50 percent in 3 patients (21 percent), and 50 to 75 percent in 2 patients (14 percent). All patients exhibited evidence of fat necrosis, and 7 patients (50 percent) also developed necrosis of a portion of the TRAM flap skin island. Additionally, mastectomy flap necrosis developed in 5 patients (36 percent). For purposes of identification, fat necrosis was local-
TRAM, transverse rectus abdominis musculocutaneous; IM, inferomedial; IL, inferolateral; C, central; SL, superolateral; SM, superomedial.
ized to the inferolateral, inferomedial, superolateral, superomedial, and central quadrants of the reconstructed breast. Of the 70 potential quadrants in the reconstructed breasts included in this study, fat necrosis developed in 25 (36 percent). Four quadrants were involved in 1 patient, three quadrants in 1, two quadrants in 6, and one quadrant in 6. The superolateral quadrant was involved in 4 instances, the superomedial in 1, the inferolateral in 10, and the inferomedial in 9 cases (Fig. 1). The superficial layer of fat was harvested with the latissimus dorsi flap to augment flap volume in 8 patients (57 percent), and the nerve was cut to prevent postoperative motion in the reconstructed breast in 3 patients (21 percent). The skin island from the latissimus dorsi flap was used to participate in the formation of the skin envelope in the reconstructed breast in 9 patients (64 percent). One patient developed a hematoma in the back donor site, and seroma formation requiring aspiration after removal of the drain was documented in 8 patients (57 percent). The time from the original TRAM to the latissimus flap secondary reconstruction ranged from 2 to 14 months, with 11 of these patients requiring secondary reconstruction within the first 12 months after their TRAM flap procedure.
All 14 latissimus dorsi secondary breast reconstructions survived completely and produced a successful reconstruction (Figs. 2 through 4). Patient satisfaction and opinion of the overall result were determined by telephone follow-up. Of the 14 patients involved in the study, 4 were unavailable for follow-up. Of the 10 patients remaining, all found the latissimus salvage of their reconstruction worthwhile. Patients also reported that recovery from the latissimus flap reconstruction was overall shorter and easier compared with recovery after the TRAM flap procedure.
DISCUSSION
The clinical scenario described in this article is a challenging problem for two reasons. First, a technique for breast reconstruction specifically based on the use of autologous tissue alone has been only partially successful in creating an aesthetic breast mound, and revision is required. To then opt for an implant-based solution introduces all of the potential complications associated with the use of breast implants. This step seems to defeat the purpose of using the TRAM flap in the first place, and many patients are reluctant to accept the use of these devices. Second, the defect that results from partial TRAM flap loss is generally irregular with sharply defined borders. Such defects are generally not smoothly “filled in” with a breast implant, and the aesthetic result can be less than ideal. On both accounts, the latissimus dorsi musculocutaneous flap affords an ideal solution to this difficult problem. Because the flap is very reliable, has sufficient volume, and transposes easily into the defect, a technique to preserve the autologous nature of the reconstruction is readily provided. As well, and perhaps more importantly, the pliable nature of the flap coupled with the vigorous blood supply allow the flap to be folded and inset into even very irregular defects without the risk of significant flap ischemia. For these reasons, we believe the use of th