BREAST Microvascular Modifications to Optimize the Transverse Upper Gracilis Flap for Breast Reconstruction Judith E. Hunter, F.R.C.S.(Plast.) Simon P. Mackey, F.R.C.S.(Plast.) Radovan Boca, Ph.D., F.R.C.S.(Plast.)Eq. Paul A. Harris, F.R.C.S.(Plast.) London, United Kingdom
Background: The free transverse upper gracilis flap is well described for breast reconstruction and is the authors’ second choice. Medial thigh soft tissue creates a durable, pliable, aesthetic breast; however, it has been criticized for modest volume and short pedicle. This demands frequent use of bilateral flaps for unilateral reconstructions, sacrifice of thoracodorsal vessels, and/or use of vein grafts. The authors have overcome these issues by modifying their microvascular techniques. Methods: The authors describe several maneuvers that they have introduced, including excision and replacement of costal cartilage, using nontraditional internal mammary arterial anastomoses, and using adductor branches for flap-to-flap anastomoses to allow double flap reconstructions. The authors describe their case series of 30 transverse upper gracilis flaps to reconstruct 20 breasts in 18 patients. Results: All flaps have survived. Seventy-five percent of the reconstructions were unilateral, although of these, 67 percent used two flaps to reconstruct one breast. The mean reconstructed breast was 360.9 g. Of 10 breasts reconstructed with double flaps, six used available adductor branches, whereas the others used internal mammary perforators, end-to-side anastomoses, or retrograde arterial flow. To complement this clinical approach, an anatomical study of the branching patterns from the gracilis pedicle has been carried out on 33 cadaveric specimens. Conclusions: The authors’ study has allowed a new classification system to be defined and demonstrates suitable branching patterns to allow flap-to-flap anastomoses in 75 percent of patients; in those 25 percent where this is not possible, alternative strategies for double flaps can be sought and have been used successfully in our clinical series. (Plast. Reconstr. Surg. 133: 1315, 2014.) CLINICAL QUESTION/LEVEL OF EVIDENCE: Therapeutic, IV.
T
he free transverse upper gracilis or transverse myocutaneous gracilis flap is well described in the literature for autologous breast reconstruction and is gaining popularity.1–11 The transverse upper gracilis flap uses upper medial thigh tissue with a transversely oriented skin paddle, From the Department of Plastic Surgery, The Royal Marsden Hospital. Received for publication April 21, 2013; accepted November 15, 2013. Presented at the American Society for Reconstructive Microsurgery 2013 Annual Meeting, in Naples, Florida, January 12 through 15, 2013; and the British Association of Plastic, Reconstructive and Aesthetic Plastic Surgeons Winter Meeting 2012, in London, United Kingdom, December 4 through 7, 2012. Copyright © 2014 by the American Society of Plastic Surgeons DOI: 10.1097/PRS.0000000000000173
skewed posteriorly. This allows maximal use of the angiosome of the dominant gracilis pedicle.1,2,11–14 Transverse upper gracilis flaps, in our experience, have a defined place among the techniques available to breast surgeons. We have used them in the delayed and immediate setting after mastectomy and in unilateral or bilateral reconstructions. They are now our flap of choice when there is insufficient pannus to undertake an abdominally based reconstruction, as there is usually enough tissue in the inner thighs of patients with even the most athletic builds to allow reconstruction of a moderate breast. We have also used them in salvage cases, after failed reconstructions with implants (with or without dermal substitutes or Disclosure: The authors have no financial interest to declare in relation to the content of this article.
www.PRSJournal.com
1315
Plastic and Reconstructive Surgery • June 2014 latissimus dorsi flaps) or deep inferior epigastric perforator (DIEP) flaps. Compared with buttock-based flaps, we have found that the transverse upper gracilis flap is more straightforward to raise, does not require turning of the patient, and provides a well-hidden scar. The flap itself has pliable adipose tissue, which is more suited than buttock tissue to breast reconstruction. For moderate breasts, it usually yields more tissue than the extended latissimus dorsi flap and does not usually result in problematic seromas, visible scarring, or asymmetry15 or rely on an often irradiated thoracodorsal pedicle. We recognize that the flap has been criticized for donor-site problems, namely, wound breakdown, poor scarring, and sensory disturbance,5,16 and for loss of volume with time, attributed to muscle atrophy.16 However, we do not rely on the muscle for bulk, taking only enough proximal muscle to allow easy harvest of the pedicle.5 Modifications including resuspension of the superficial fascial system and preservation of the long saphenous vein have been well described to address these issues.11 Other shortcomings of the flap include the short pedicle and modest volume. Since the introduction of free transverse upper gracilis flaps to our surgical repertoire for autologous breast reconstruction at The Royal Marsden Hospital in 2010, we have made several technical refinements to overcome these problems, which we have not found reported elsewhere. This article describes these microvascular modifications and offers a review of our case series. The anatomical basis required for these changes was then investigated with a cadaveric study, leading to a proposed classification that aids intraoperative decision making.
PATIENTS AND METHODS Technical Microvascular Descriptions and Case Series The process of raising the transverse upper gracilis flap is well described elsewhere.5,11 In this article, we have focussed on the microvascular modifications we have made to overcome problems with short pedicle length and modest flap volume. Approaches to Overcoming the Short Gracilis Pedicle: Optimizing Internal Mammary Vessel Length The gracilis has a short major pedicle, measured in cadaveric studies to be a mean length of 6.7 cm.14 To maximize this length, we not only
1316
retract the adductor longus to harvest the pedicle as close to its origin from the profunda femoris5,17 but also prepare long lengths of recipient vessels. We routinely adopt a rib-sparing technique to preserve chest contour when preparing the internal mammary vessels for DIEP flap reconstruction, and have modified this technique further to compensate for the relatively short gracilis pedicle. We prepare a longer length of recipient vessel by temporarily excising one of the costal cartilages (typically the third or fourth) in the subperichondrial plane, dissecting out the internal mammary vessels beneath and in the interspaces above and below. The vessels are then divided distally and retracted superiorly to allow the rib cartilage to be replaced into its perichondrial lining, repaired using 4-0 Vicryl sutures (Ethicon, Inc., Somerville, N.J.). The recipient vessels are then laid over this repair, creating a long recipient pedicle (Fig. 1). This allows ample space, bringing the vessels out of the depths of the breast pocket. Tension-free anastomoses can then be performed while preserving chest contour. We have not had to resort to vein grafts using this technique. Obtaining Adequate Volume Using Double Flaps Centering the flap more posteriorly (taking advantage of the availability of tissue here and the angiosome of the gracilis vessels) and harvesting fat beyond the skin paddle are all well-recognized techniques to maximize tissue from the inner thigh. We have harvested flaps between 161 and 500 g (Table 1), similar to that reported in the literature.16 To overcome this relatively small volume, we frequently harvest bilateral flaps to reconstruct single breasts (double flaps). Our mean reconstructed breast is therefore 361 g. This, however, presents challenges in terms of availability of recipient vessels; other authors have combined internal mammary and thoracodorsal recipients with vein grafts to allow freedom of inset. We have a preferred approach to this, performing flap-to-flap anastomoses, moving two flaps en bloc or “stacked” up to the chest, necessitating only a single arterial anastomosis with the internal mammaries, similar to the “mosaic” principle introduced by Koshima et al.18 We have often observed a large branch from the main gracilis pedicle to the adductor muscles, comparable in size to the pedicle itself. We have made use of these branches, joining them to the main pedicle of the second flap, allowing both flaps to be powered by the internal mammary vessels (Fig. 2). For ease of flap positioning and to minimize ischemic time, we first anastomose
Volume 133, Number 6 • Transverse Upper Gracilis Flap
Fig. 1. (Above, left) Intraoperative photograph of breast pocket; costal cartilage is removed from the perichondrium and reflected to expose the internal mammary vessels. (Above, right) Internal mammary vessels are divided distally and reflected superiorly, and costal cartilage is then replaced. (Below, left) Long length of internal mammary vessels placed over costal cartilage. (Below, right) Comfortable anastomosis of transverse upper gracilis flap.
one flap to the internal mammaries and then join the second flap in series by means of the adductor branch. To our knowledge, this has not been described before in the literature. When this is not possible, we have successfully anastomosed one transverse upper gracilis flap end to side and one end to end on a single internal mammary artery, one transverse upper gracilis flap end to end to the internal mammary vessels and one end to end on a suitable internal mammary perforator, and one transverse upper gracilis flap anterograde and one retrograde end
to end using a single internal mammary artery (Fig. 3). We were happy to perform retrograde arterial anastomoses, as flow from the distal internal mammary artery appeared good. This we have measured quantitatively as part of another study using arterial line transducers. In one of our cases using this method for bilateral transverse upper gracilis flaps, the mean arterial pressure was 44 mmHg in the distal internal mammary artery compared with 60 mmHg in the proximal end, whereas the systemic mean arterial pressure was 70 mmHg.
1317
Plastic and Reconstructive Surgery • June 2014 Table 1. Patient and Flap Characteristics Characteristic
Value
Age, yr Mean Range Body mass index, kg/m2 Mean Range Smokers, % Chemotherapy, % Radiotherapy, % No. of breasts undergoing immediate reconstruction No. of breasts undergoing delayed reconstruction No. of breasts undergoing salvage reconstruction No. of breasts undergoing unilateral reconstruction No. of breasts undergoing bilateral reconstruction No. of bilateral flaps for unilateral reconstruction Costal cartilage excised and replaced (no. of breasts) Mastectomy mass in immediate reconstructions (n = 4), g Mean Range Mass of flap, g Mean Range Mass of reconstructed breast, g Mean Range Ischemic time, min Mean Range
51.6 39–67 22.9 19.1–31.1 0 56 50 4 (20%) 16 (80%) 4 (25% of delayed reconstructions) 15 (75% reconstructions) 5 (25% reconstructions) 10 (67% unilateral reconstructions) 17 (85% reconstructions) 134 110–158 244.6 161–500 360.9 165–711 52.6 35–71
Case Series We have applied these methods to 30 transverse upper gracilis flaps, reconstructing 20 breasts in 18 patients. All flaps were anastomosed to the internal mammary vessels without use of vein grafts or sacrifice of the thoracodorsal vessels. All flaps have survived. As shown in Table 1, the mean age of our patients was 51.6 years with, as expected, a mean body mass index at the lower end of the spectrum (22.9). None of our patients were smokers, although approximately half had been subjected to preoperative chemotherapy or radiotherapy. The majority of cases were delayed reconstructions, with 25 percent of these representing salvage procedures for previous failed reconstructive attempts (including implant only, latissimus dorsi and implant, and DIEP flap reconstructions). Seventy-five percent of patients had unilateral reconstructions although, of these, 67 percent used two flaps to reconstruct one breast (double flaps). Our mean reconstructed breast was therefore 360.9 g, which represents approximately 60 percent of our mean DIEP flap mass; this is respectable, given the slighter nature of the cohort appropriate for transverse upper gracilis compared with DIEP flap reconstruction. Of our 10 breasts reconstructed with double flaps, available adductor branches were used in six (Fig. 4), and internal mammary perforators, end-to-side
Fig. 2. (Left) The first flap anastomosed to the internal mammary vessels. The branch from the main pedicle is shown hanging down dependently. (Right) The second flap is anastomosed by means of a branch from the gracilis pedicle.
1318
Volume 133, Number 6 • Transverse Upper Gracilis Flap
Fig. 3. Diagrams of double flap anastomotic options using internal mammary axis. (Above, left) Flap-to-flap by means of the adductor branch. (Above, right) One flap end to side and one end to end on the internal mammary artery. (Below, left) One flap on the internal mammary perforator and one on the internal mammary vessels proper. (Below, right) One flap connected retrograde and one anterograde on the internal mammary artery.
anastomoses, or retrograde flow was used in the others. Of the eight cases where we searched for suitable gracilis pedicle branches, two had no useful branches on either leg (25 percent). Of those cases where suitable branches were found, the majority used were branches to the adductor magnus (n = 5), with an adductor longus branch being used in only one case. Our mean ischemic time was 52.6 minutes, and in 17 of our breast reconstructions (85 percent), we excised and replaced costal cartilage to gain compensatory recipient pedicle length. Our complications included one return to the operating room for early flap compromise requiring reinset of the flap (one in 20 breasts). Six donor
sites have proved problematic (six of 30 legs), with either ongoing seromas, pain, or wound dehiscence. Two thighs required reoperation and two thighs (one patient) required prolonged vacuum-assisted closure. Cadaveric Study Our novel use of the branches from the gracilis pedicle to enable flap-to-flap anastomoses and therefore facilitate double flaps led us to ponder the variability of the branching patterns. As we increasingly used this technique, it was noticed that some patients had several potential branches that could be used for vascular anastomoses, whereas others had very few. If we could
1319
Plastic and Reconstructive Surgery • June 2014
Fig. 4. Double flap reconstruction following implant removal using flap-to-flap technique. (Left) Preoperatively; (right) postoperatively.
predict where we might most reliably find these branches and even be forewarned of a paucity of them, this would speed up our flap harvest and enable better planning. The vascular pedicle to the gracilis, its length and origin, and its cutaneous perforator territory have already been well described.1,12–14,17,19–21 The branches from the pedicle have been described in cadaveric and clinical articles as requiring division to raise the standard flap.17,22 However, their variability and utility for flap-to-flap anastomoses have not been investigated for breast reconstruction. One cadaveric dissection article from the Asian population has looked at the feasibility of taking double functional muscle free flaps of gracilis and adductor longus using a single pedicle and has classified the pedicles according to pattern and length.23 They did not, however, specifically investigate the size of these branches and alternative branches to other muscles (such as adductor brevis or longus), which would be as efficacious as a vascular conduit for flap-to-flap designs for breast reconstruction. Method Cadaveric soft fixed specimens were studied. The dominant gracilis pedicle was identified in an atraumatic fashion and both artery and veins were dissected back to their origin, typically from the profunda vessels. Any branches were then preserved, their destination was noted, and all vessels were measured in length
1320
and diameter using digital callipers (Maplin Electronics, South Yorkshire, United Kingdom; resolution, 0.01 mm; accuracy, ±0.02 mm). All measurements were obtained by a single investigator (R.B.) and noted simultaneously by a second investigator (J.H.). Where two veins were accompanying an artery, the largest was measured. For branches to be deemed feasible for flap-to-flap anastomoses, they had to be at least 10 mm in length, at least 10 mm away from the gracilis muscle or the origin of the pedicle, and at least 1 mm in diameter to facilitate imagined use of clamps and venous coupler device in vivo. The data were recorded in an Excel (Microsoft Corp., Redmond, Wash.) database and a classification system was proposed.
RESULTS Thirty-three thighs were dissected in 19 female and 14 male specimens; 17 were left thighs and 16 were right thighs. The mean main pedicle length was 6.83 cm (range, 4.32 to 11.95 cm), with a mean arterial diameter of 2.83 mm (range, 1.9 to 3.8 mm) and a mean venous diameter of 3.24 mm (range, 1.5 to 5 mm). All specimens had some muscular branches coming from the main pedicle, but using our cutoffs for feasibility of microvascular anastomoses, we classified the patterns as either being G0, GB, GL, or GM, or a combination where several usable branches were found to different muscles. G0 was defined as having no usable branches,
Volume 133, Number 6 • Transverse Upper Gracilis Flap GB was defined as having a usable branch to the adductor brevis muscle, GL was defined as having a usable branch to the adductor longus muscle, and GM was defined as having a usable branch to the adductor longus (Fig. 5). Using our classification, 22 thighs had suitable branches for flap-to-flap anastomosis, whereas 11 did not (classified therefore as G0). An example of a GM dissection is shown in Figure 6. Of those with suitable branching patterns, four were GB, six were GL, and eight were GM. A further four thighs had more than one suitable branch (GB/GL or GB/GL/GM) (Fig. 7). Of those with usable branches, the vessel diameters and lengths are as shown in Table 2. In those specimens where both thighs were dissected (n = 24), the symmetry between sides was examined. Seven of 12 cadavers displayed symmetry in vascular branching, whereas five did not. This encompassed the full range of branching patterns; thus, symmetry appeared no more likely for G0, GM, GB, or GL thighs. If one side was G0 in these specimens (n = 10), the other side was G0 in three, GM in two, and GB in two. This means that, although in 75 percent of patients suitable branches could be found to
Fig. 5. Schematic of branches from the main gracilis pedicle to surrounding muscles. GL, usable branch to the adductor longus; GB, usable branch to the adductor brevis; GM, usable branch to the adductor magnus.
Fig. 6. Anatomical image of gracilis pedicle with large adductor magnus branch.
allow flap-to-flap anastomoses, correspondingly, in 25 percent, potentially no good branches would be available.
DISCUSSION Compared with the more established DIEP flap for breast reconstruction, there are few series and technical tips published for free transverse upper gracilis flaps. Series published range from single figures to 154, with most numbering in the teens.1–9,24 Our series adds to the literature, but is primarily a descriptive microsurgical technique and anatomical dissection article. Similar to large series of transverse upper gracilis flaps used for breast reconstruction, we report few complications with this flap and good cosmetic results.7,9,15 Our first few patients did have problems with seromas, wound breakdown, and neuropathic pain at the donor site.24,25 These issues now occur more rarely because of, we think, resuspending the superficial fascial system to the deep fascia in the groin and periosteum, keeping drains in the thigh for up to 3 weeks (patients are educated to manage their drains at home after discharge from the hospital), avoiding harvest of the long saphenous vein with the flap,25 carefully preserving the posterior cutaneous branches of the femoral nerve, and being conservative with skin paddle dimensions. This has become more possible as we continue to improve our double flap techniques, so that flap volume and skin paddle size can be maximized without having tight donor-site closure. Excellent cosmetic results can be obtained from this donor site (Fig. 8). Our microvascular modifications have evolved with time and experience. We have had no flap failures in our series of 30 flaps, which
1321
Plastic and Reconstructive Surgery • June 2014
Fig. 7. Chart showing variation in anatomical patterns among cadaveric specimens. G0, no usable branches; GB, usable branch to the adductor brevis; GL, usable branch to the adductor longus; GM, usable branch to the adductor magnus (see text for details).
Table 2. Table of Cadaveric Measurements of Vessel Branches that Fulfilled Criteria as Being Usable for Microvascular Anastomosis Branch Length (cm)
Branch Artery Diameter (mm)
Branch Vein Diameter (mm)
Type
Mean
Range
Mean
Range
Mean
Range
GB GL GM
2.98 1.37 2.57
1.19–5.72 1.03–2.05 1.1–5.75
1.66 1.69 2.02
1.21–2.55 1.09–2.50 1.09–2.70
1.91 2.13 2.04
1.08–3.28 1.05–4.10 1.29–2.60
GB, usable branch to the adductor brevis; GL, usable branch to the adductor longus; GM, usable branch to the adductor magnus.
Fig. 8. Postoperative donor sites. (Left) Anterior and (right) posterior.
perhaps reflects a mature microsurgical practice, with approximately 250 free flaps being performed in total in our unit per year. All flaps have achieved their primary aim of breast
1322
reconstruction, without recourse to supplementary flaps or implants, in contrast to a recent series quoting reconstructive failure in onethird of free transverse upper gracilis flaps.16
Volume 133, Number 6 • Transverse Upper Gracilis Flap Nine percent of the patients in the study by Buntic et al. required further augmentation with implants.24 However, we do fairly routinely perform lipofilling and/or contralateral balancing surgery as a secondary procedure to finesse the breast mound, but this is true also of our DIEP flap reconstructions. In contrast to the approach suggested by other authors,11 when bilateral stacked transverse upper gracilis flaps are used to recreate one breast mound, we prefer to use the internal mammary axis for both flaps rather than the thoracodorsal vessels in addition, as this simplifies dissection, anastomoses, and inset. Like others, we did initially use the thoracodorsal vessels or serratus branches for second pedicle anastomoses but struggled to medialize the flap without the use of vein grafts and returned a patient to the operating room because of a hematoma following early inadvertent abduction of the patient’s arm. We therefore sought these alternative approaches using the internal mammary recipients and have found these to work well. Since adopting solely the internal mammary axis, we have returned only one patient immediately to the operating room from recovery because of a problem with inset causing kinking of the vessels. This was rectified easily and we have had no flap losses. We have not needed to use vein grafts in any of our transverse upper gracilis flaps, in contrast to other series.8 End-to-side and end-to-end anastomoses to the internal mammary vessels and use of both retrograde and anterograde arterial flow are maneuvers we have performed in our series, procedures mentioned but not described in detail in other series.25 We now favor use of flap-to-flap anastomosis with an adductor branch, a technique we have not seen published elsewhere. Perhaps because of our comfort with stacking our flaps in series in this way, a large proportion (67 percent) of our patients have had bilateral flaps for reconstructing unilateral defects, which enables a mass of tissue to be transferred to rival DIEP flaps. This avoids using a vertical scar in the medial thigh used by some authors to create volume.25,26 Other series quote smaller proportions of double flaps (15 percent8 and 21 percent11). Rather than take a very large single flap, with possible adverse donor-site sequelae such as wound dehiscence or asymmetry, we have a low threshold for using these bilateral flaps. Our anatomical study was prompted because there were no previous studies looking at the precise branching patterns of the gracilis pedicle
for us to refer to. We aimed therefore to investigate whether suitable branches would universally be found, whether there was symmetry between thighs of the same patient, where the most reliable branches would likely be situated, and as a consequence whether preoperative imaging was likely to be beneficial. In our study, the mean overall length and diameter of the main pedicle was similar to that previously published,14,23 which lends weight to the accuracy of our measurement. Certainly, our methods are standard in basic anatomical research, and inaccuracies were minimized by a single investigator measuring all of the parameters. We felt our cutoffs of a minimum of 1 cm for vessel branch length and 1 mm for vessel diameter (both artery and vein were required to fulfil these criteria) were reasonable for our microsurgical practice. Of those thighs that had suitable vessels for flap-to-flap anastomosis, eight had GB branches, had 10 GL branches, and nine had GM branches. However, the GB and GM branches were generally longer than the GL ones, and the vessel diameters were larger for the GM pattern. This is fortunate, as branches to the adductor magnus are more accessible than those to the adductor longus or brevis in vivo. The only study we could find that looked at the branches from the gracilis pedicle specifically investigated the possibility of transplanting both gracilis and adductor longus muscles with a single set of anastomoses.23 This was viewing the situation essentially in reverse and did not include lengths or diameters of the branches themselves or details of branches to other surrounding muscles. They also showed considerable variability, with 46 percent showing single branches to each muscle and 50 percent having multiple branches. In addition, they described short and ultrashort common pedicle branches very close to the profunda vessels, but this would also cause some problems with our inset. In our dissections where there was a GL pattern, the length from the profunda vessels to this branch was a mean of 2.97 cm (minimum, 1.19 cm), which is similar to that recorded by Sananpanich et al. in 2008. However, this length was, as expected, considerably longer in the GM specimens (mean, 4.87 cm) and shorter in the GB specimens (mean, 1.22 cm). Interestingly, 34 percent of the cadaveric thighs were G0 and therefore in vivo would not have been suitable for flap-to-flap anastomosis. However, 66 percent would have been, and of those paired thighs from single cadavers, having a G0 pattern on one side was not necessarily predictive of a G0
1323
Plastic and Reconstructive Surgery • June 2014 pattern contralaterally. Only one of the two thighs of course is required to have a suitable branching pattern for flap-to-flap anastomosis to be possible. Therefore, in these 12 cadavers with both thighs available for dissection, three would not have had suitable branches on either side (25 percent). Although these numbers are small, we could therefore estimate that 25 percent of our patients will not have suitable branches and 75 percent will. In the cases where no suitable branch is found on either side, double flaps are still possible, but alternative strategies need to be found. In our clinical series, we have encountered this also in 25 percent of our patient cohort where branches were sought, and have still managed to avoid use of the thoracodorsal vessels or vein grafts by using perforator vessels or end-to-side or retrograde anastomoses. However, it would be useful to know preoperatively which leg has suitable branches to allow more efficient flap harvest. In addition, to be forewarned of the lack of suitable branches being available on either thigh would allow the surgeon to ensure that any perforating vessels from the internal mammaries are preserved, or in the case of the requirement for retrograde anastomosis, the internal mammary vessels can be divided at the center of the dissected space rather than divided too far distally (Fig. 9). To this end, we have started to arrange preoperative imaging for our future transverse upper gracilis flap reconstruction patients. Eventually, we may evolve to using profunda artery perforator flaps27
in selected patients with suitable anatomy for a supine harvest, with the transverse upper gracilis flap as a good option in those with very posterior perforators on imaging.
CONCLUSIONS The transverse upper gracilis flap is now our second choice for autologous reconstruction of the breast when the abdomen is not available or insufficient. It does present the reconstructive surgeon with challenges, particularly in obtaining adequate volume and comfortable anastomosis given a short pedicle. This article outlines some of the novel microvascular approaches we have used to overcome these problems. By use of intraoperative observation of vascular anatomy and physiology, we have successfully adapted our standard techniques. We have optimized recipient pedicle length by means of costal cartilage removal and replacement, and performed nontraditional internal mammary and flap-to-flap anastomoses to allow double flaps to be used. The availability of branches from the gracilis pedicle to allow flap-to-flap anastomoses has not specifically been investigated before, and we present our cadaveric anatomical findings and a proposed classification system of these branching patterns. Knowledge of this detailed anatomy may allow more efficient flap raising and anastomotic planning; preoperative imaging may also be indicated.
Fig. 9. Decision-making algorithm for double transverse upper gracilis flap reconstruction. TUG, transverse upper gracilis flap; IM, internal mammary; IMA, internal mammary artery.
1324
Volume 133, Number 6 • Transverse Upper Gracilis Flap Judith E. Hunter, F.R.C.S.(Plast.) The Royal Marsden Hospital Fulham Road London SW3 6JJ, United Kingdom [email protected]
REFERENCES 1. Yousif NJ. The transverse gracilis musculocutaneous flap. Ann Plast Surg. 1993;31:382. 2. Wechselberger G, Schoeller T, Bauer T, et al. Surgical technique and clinical application of the transverse gracilis myocutaneous free flap. Br J Plast Surg. 2001;54:423–427. 3. Wechselberger G, Schoeller T. The transverse myocutaneous gracilis free flap: A valuable tissue source in autologous breast reconstruction. Plast Reconstr Surg. 2004;114:69–73. 4. Schoeller T, Huemer GM, Kolehmainen M, Otto-Schoeller A, Wechselberger G. A new “Siamese” flap for breast reconstruction: The combined infragluteal-transverse myocutaneous gracilis muscle flap. Plast Reconstr Surg. 2005;115:1110–1117. 5. Arnez ZM, Pogorelec D, Planinsek F, Ahcan U. Breast reconstruction by the free transverse gracilis (TUG) flap. Br J Plast Surg. 2004;57:20–26. 6. Schoeller T, Wechselberger G. Breast reconstruction by the free transverse gracilis (TUG) flap. Br J Plast Surg. 2004;57:481–482. 7. Fansa H, Schirmer S, Warnecke IC, Cervelli A, Frerichs O. The transverse myocutaneous gracilis muscle flap: A fast and reliable method for breast reconstruction. Plast Reconstr Surg. 2008;122:1326–1333. 8. Schoeller T, Huemer GM, Wechselberger G. The transverse musculocutaneous gracilis flap for breast reconstruction: Guidelines for flap and patient selection. Plast Reconstr Surg. 2008;122:29–38. 9. Vega SJ, Sandeen SN, Bossert RP, Perrone A, Ortiz L, Herrera H. Gracilis myocutaneous free flap in autologous breast reconstruction. Plast Reconstr Surg. 2009;124:1400–1409. 10. McCulley SJ, Macmillan RD, Rasheed T. Transverse upper gracilis (TUG) flap for volume replacement in breast conserving surgery for medial breast tumours in small to medium sized breasts. J Plast Reconstr Aesthet Surg. 2011;64:1056–1060. 11. Fattah A, Figus A, Mathur B, Ramakrishnan VV. The transverse myocutaneous gracilis flap: Technical refinements. J Plast Reconstr Aesthet Surg. 2010;63:305–313. 12. Yousif NJ, Matloub HS, Kolachalam R, Grunert BK, Sanger JR. The transverse gracilis musculocutaneous flap. Ann Plast Surg. 1992;29:482–490. 13. Coquerel-Beghin D, Milliez PY, Auquit-Auckbur I, Lemierre G, Duparc F. The gracilis musculocutaneous flap: Vascular supply of the muscle and skin components. Surg Radiol Anat. 2006;28:588–595.
14. Wong C, Mojallal A, Bailey SH, Trussler A, Saint-Cyr M. The extended transverse musculocutaneous gracilis flap: Vascular anatomy and clinical implications. Ann Plast Surg. 2011;67:170–177. 15. Pülzl P, Schoeller T, Kleewein K, Wechselberger G. Donorsite morbidity of the transverse musculocutaneous gracilis flap in autologous breast reconstruction: Short-term and long-term results. Plast Reconstr Surg. 2011;128:233e–242e. 16. Locke MB, Zhong T, Mureau MA, Hofer SO. Tug ‘O’ war: Challenges of transverse upper gracilis (TUG) myocutaneous free flap breast reconstruction. J Plast Reconstr Aesthet Surg. 2012;65:1041–1050. 17. Hasen KV, Gallegos ML, Dumanian GA. Extended approach to the vascular pedicle of the gracilis muscle flap: Anatomical and clinical study. Plast Reconstr Surg. 2003;111:2203–2208. 18. Koshima I, Yamamoto H, Moriguchi T, Orita Y. Extended anterior thigh flaps for repair of massive cervical defects involving pharyngoesophagus and skin: An introduction to the “mosaic” flap principle. Ann Plast Surg. 1994;32:321–327. 19. An X, Yue B, Lee JH, Lin C, Han SH. Arterial anatomy of the gracilis muscle as determined by latex injection and glycerin transparency. Clin Anat. 2012;25:231–234. 20. Lykoudis EG, Spyropoulou GA, Vlastou CC. The ana tomic basis of the gracilis perforator flap. Br J Plast Surg. 2005;58:1090–1094. 21. Whitaker IS, Karavias M, Shayan R, et al. The gracilis myocutaneous free flap: A quantitative analysis of the fasciocutaneous blood supply and implications for autologous breast reconstruction. PLoS One 2012;7:e36367. 22. Hallock GG. The gracilis (medial circumflex femoral) perforator flap: A medial groin free flap? Ann Plast Surg. 2003;51:623–626. 23. Sananpanich K, Tu YK, Pookhang S, Chalidapong P. Anatomic variance in common vascular pedicle of the gracilis and adductor longus muscles: Feasibility of double functioning free muscle transplantation with single pedicle anastomosis. J Reconstr Microsurg. 2008;24:231–238. 24. Buntic RF, Horton KM, Brooks D, Althubaiti GA. Transverse upper gracilis flap as an alternative to abdominal tissue breast reconstruction: Technique and modifications. Plast Reconstr Surg. 2011;128:607e–613e. 25. Saint-Cyr M, Wong C, Oni G, et al. Modifications to extend the transverse upper gracilis flap in breast reconstruction: Clinical series and results. Plast Reconstr Surg. 2012;129:24e–36e. 26. Kind GM, Foster RD. The longitudinal gracilis myocuta neous flap: Broadening options in breast reconstruction. Ann Plast Surg. 2008;61:513–520. 27. Allen RJ, Haddock NT, Ahn CY, Sadeghi A. Breast reconstruction with the profunda artery perforator flap. Plast Reconstr Surg. 2012;129:16e–23e.
1325
Background: The free transverse upper gracilis flap is well described for breast reconstruction and is the authors’ second choice. Medial thigh soft tissue creates a durable, pliable, aesthetic breast; however, it has been criticized for modest volume and short pedicle. This demands frequent use of bilateral flaps for unilateral reconstructions, sacrifice of thoracodorsal vessels, and/or use of vein grafts. The authors have overcome these issues by modifying their microvascular techniques. Methods: The authors describe several maneuvers that they have introduced, including excision and replacement of costal cartilage, using nontraditional internal mammary arterial anastomoses, and using adductor branches for flap-to-flap anastomoses to allow double flap reconstructions. The authors describe their case series of 30 transverse upper gracilis flaps to reconstruct 20 breasts in 18 patients. Results: All flaps have survived. Seventy-five percent of the reconstructions were unilateral, although of these, 67 percent used two flaps to reconstruct one breast. The mean reconstructed breast was 360.9 g. Of 10 breasts reconstructed with double flaps, six used available adductor branches, whereas the others used internal mammary perforators, end-to-side anastomoses, or retrograde arterial flow. To complement this clinical approach, an anatomical study of the branching patterns from the gracilis pedicle has been carried out on 33 cadaveric specimens. Conclusions: The authors’ study has allowed a new classification system to be defined and demonstrates suitable branching patterns to allow flap-to-flap anastomoses in 75 percent of patients; in those 25 percent where this is not possible, alternative strategies for double flaps can be sought and have been used successfully in our clinical series. (Plast. Reconstr. Surg. 133: 1315, 2014.) CLINICAL QUESTION/LEVEL OF EVIDENCE: Therapeutic, IV.
T
he free transverse upper gracilis or transverse myocutaneous gracilis flap is well described in the literature for autologous breast reconstruction and is gaining popularity.1–11 The transverse upper gracilis flap uses upper medial thigh tissue with a transversely oriented skin paddle, From the Department of Plastic Surgery, The Royal Marsden Hospital. Received for publication April 21, 2013; accepted November 15, 2013. Presented at the American Society for Reconstructive Microsurgery 2013 Annual Meeting, in Naples, Florida, January 12 through 15, 2013; and the British Association of Plastic, Reconstructive and Aesthetic Plastic Surgeons Winter Meeting 2012, in London, United Kingdom, December 4 through 7, 2012. Copyright © 2014 by the American Society of Plastic Surgeons DOI: 10.1097/PRS.0000000000000173
skewed posteriorly. This allows maximal use of the angiosome of the dominant gracilis pedicle.1,2,11–14 Transverse upper gracilis flaps, in our experience, have a defined place among the techniques available to breast surgeons. We have used them in the delayed and immediate setting after mastectomy and in unilateral or bilateral reconstructions. They are now our flap of choice when there is insufficient pannus to undertake an abdominally based reconstruction, as there is usually enough tissue in the inner thighs of patients with even the most athletic builds to allow reconstruction of a moderate breast. We have also used them in salvage cases, after failed reconstructions with implants (with or without dermal substitutes or Disclosure: The authors have no financial interest to declare in relation to the content of this article.
www.PRSJournal.com
1315
Plastic and Reconstructive Surgery • June 2014 latissimus dorsi flaps) or deep inferior epigastric perforator (DIEP) flaps. Compared with buttock-based flaps, we have found that the transverse upper gracilis flap is more straightforward to raise, does not require turning of the patient, and provides a well-hidden scar. The flap itself has pliable adipose tissue, which is more suited than buttock tissue to breast reconstruction. For moderate breasts, it usually yields more tissue than the extended latissimus dorsi flap and does not usually result in problematic seromas, visible scarring, or asymmetry15 or rely on an often irradiated thoracodorsal pedicle. We recognize that the flap has been criticized for donor-site problems, namely, wound breakdown, poor scarring, and sensory disturbance,5,16 and for loss of volume with time, attributed to muscle atrophy.16 However, we do not rely on the muscle for bulk, taking only enough proximal muscle to allow easy harvest of the pedicle.5 Modifications including resuspension of the superficial fascial system and preservation of the long saphenous vein have been well described to address these issues.11 Other shortcomings of the flap include the short pedicle and modest volume. Since the introduction of free transverse upper gracilis flaps to our surgical repertoire for autologous breast reconstruction at The Royal Marsden Hospital in 2010, we have made several technical refinements to overcome these problems, which we have not found reported elsewhere. This article describes these microvascular modifications and offers a review of our case series. The anatomical basis required for these changes was then investigated with a cadaveric study, leading to a proposed classification that aids intraoperative decision making.
PATIENTS AND METHODS Technical Microvascular Descriptions and Case Series The process of raising the transverse upper gracilis flap is well described elsewhere.5,11 In this article, we have focussed on the microvascular modifications we have made to overcome problems with short pedicle length and modest flap volume. Approaches to Overcoming the Short Gracilis Pedicle: Optimizing Internal Mammary Vessel Length The gracilis has a short major pedicle, measured in cadaveric studies to be a mean length of 6.7 cm.14 To maximize this length, we not only
1316
retract the adductor longus to harvest the pedicle as close to its origin from the profunda femoris5,17 but also prepare long lengths of recipient vessels. We routinely adopt a rib-sparing technique to preserve chest contour when preparing the internal mammary vessels for DIEP flap reconstruction, and have modified this technique further to compensate for the relatively short gracilis pedicle. We prepare a longer length of recipient vessel by temporarily excising one of the costal cartilages (typically the third or fourth) in the subperichondrial plane, dissecting out the internal mammary vessels beneath and in the interspaces above and below. The vessels are then divided distally and retracted superiorly to allow the rib cartilage to be replaced into its perichondrial lining, repaired using 4-0 Vicryl sutures (Ethicon, Inc., Somerville, N.J.). The recipient vessels are then laid over this repair, creating a long recipient pedicle (Fig. 1). This allows ample space, bringing the vessels out of the depths of the breast pocket. Tension-free anastomoses can then be performed while preserving chest contour. We have not had to resort to vein grafts using this technique. Obtaining Adequate Volume Using Double Flaps Centering the flap more posteriorly (taking advantage of the availability of tissue here and the angiosome of the gracilis vessels) and harvesting fat beyond the skin paddle are all well-recognized techniques to maximize tissue from the inner thigh. We have harvested flaps between 161 and 500 g (Table 1), similar to that reported in the literature.16 To overcome this relatively small volume, we frequently harvest bilateral flaps to reconstruct single breasts (double flaps). Our mean reconstructed breast is therefore 361 g. This, however, presents challenges in terms of availability of recipient vessels; other authors have combined internal mammary and thoracodorsal recipients with vein grafts to allow freedom of inset. We have a preferred approach to this, performing flap-to-flap anastomoses, moving two flaps en bloc or “stacked” up to the chest, necessitating only a single arterial anastomosis with the internal mammaries, similar to the “mosaic” principle introduced by Koshima et al.18 We have often observed a large branch from the main gracilis pedicle to the adductor muscles, comparable in size to the pedicle itself. We have made use of these branches, joining them to the main pedicle of the second flap, allowing both flaps to be powered by the internal mammary vessels (Fig. 2). For ease of flap positioning and to minimize ischemic time, we first anastomose
Volume 133, Number 6 • Transverse Upper Gracilis Flap
Fig. 1. (Above, left) Intraoperative photograph of breast pocket; costal cartilage is removed from the perichondrium and reflected to expose the internal mammary vessels. (Above, right) Internal mammary vessels are divided distally and reflected superiorly, and costal cartilage is then replaced. (Below, left) Long length of internal mammary vessels placed over costal cartilage. (Below, right) Comfortable anastomosis of transverse upper gracilis flap.
one flap to the internal mammaries and then join the second flap in series by means of the adductor branch. To our knowledge, this has not been described before in the literature. When this is not possible, we have successfully anastomosed one transverse upper gracilis flap end to side and one end to end on a single internal mammary artery, one transverse upper gracilis flap end to end to the internal mammary vessels and one end to end on a suitable internal mammary perforator, and one transverse upper gracilis flap anterograde and one retrograde end
to end using a single internal mammary artery (Fig. 3). We were happy to perform retrograde arterial anastomoses, as flow from the distal internal mammary artery appeared good. This we have measured quantitatively as part of another study using arterial line transducers. In one of our cases using this method for bilateral transverse upper gracilis flaps, the mean arterial pressure was 44 mmHg in the distal internal mammary artery compared with 60 mmHg in the proximal end, whereas the systemic mean arterial pressure was 70 mmHg.
1317
Plastic and Reconstructive Surgery • June 2014 Table 1. Patient and Flap Characteristics Characteristic
Value
Age, yr Mean Range Body mass index, kg/m2 Mean Range Smokers, % Chemotherapy, % Radiotherapy, % No. of breasts undergoing immediate reconstruction No. of breasts undergoing delayed reconstruction No. of breasts undergoing salvage reconstruction No. of breasts undergoing unilateral reconstruction No. of breasts undergoing bilateral reconstruction No. of bilateral flaps for unilateral reconstruction Costal cartilage excised and replaced (no. of breasts) Mastectomy mass in immediate reconstructions (n = 4), g Mean Range Mass of flap, g Mean Range Mass of reconstructed breast, g Mean Range Ischemic time, min Mean Range
51.6 39–67 22.9 19.1–31.1 0 56 50 4 (20%) 16 (80%) 4 (25% of delayed reconstructions) 15 (75% reconstructions) 5 (25% reconstructions) 10 (67% unilateral reconstructions) 17 (85% reconstructions) 134 110–158 244.6 161–500 360.9 165–711 52.6 35–71
Case Series We have applied these methods to 30 transverse upper gracilis flaps, reconstructing 20 breasts in 18 patients. All flaps were anastomosed to the internal mammary vessels without use of vein grafts or sacrifice of the thoracodorsal vessels. All flaps have survived. As shown in Table 1, the mean age of our patients was 51.6 years with, as expected, a mean body mass index at the lower end of the spectrum (22.9). None of our patients were smokers, although approximately half had been subjected to preoperative chemotherapy or radiotherapy. The majority of cases were delayed reconstructions, with 25 percent of these representing salvage procedures for previous failed reconstructive attempts (including implant only, latissimus dorsi and implant, and DIEP flap reconstructions). Seventy-five percent of patients had unilateral reconstructions although, of these, 67 percent used two flaps to reconstruct one breast (double flaps). Our mean reconstructed breast was therefore 360.9 g, which represents approximately 60 percent of our mean DIEP flap mass; this is respectable, given the slighter nature of the cohort appropriate for transverse upper gracilis compared with DIEP flap reconstruction. Of our 10 breasts reconstructed with double flaps, available adductor branches were used in six (Fig. 4), and internal mammary perforators, end-to-side
Fig. 2. (Left) The first flap anastomosed to the internal mammary vessels. The branch from the main pedicle is shown hanging down dependently. (Right) The second flap is anastomosed by means of a branch from the gracilis pedicle.
1318
Volume 133, Number 6 • Transverse Upper Gracilis Flap
Fig. 3. Diagrams of double flap anastomotic options using internal mammary axis. (Above, left) Flap-to-flap by means of the adductor branch. (Above, right) One flap end to side and one end to end on the internal mammary artery. (Below, left) One flap on the internal mammary perforator and one on the internal mammary vessels proper. (Below, right) One flap connected retrograde and one anterograde on the internal mammary artery.
anastomoses, or retrograde flow was used in the others. Of the eight cases where we searched for suitable gracilis pedicle branches, two had no useful branches on either leg (25 percent). Of those cases where suitable branches were found, the majority used were branches to the adductor magnus (n = 5), with an adductor longus branch being used in only one case. Our mean ischemic time was 52.6 minutes, and in 17 of our breast reconstructions (85 percent), we excised and replaced costal cartilage to gain compensatory recipient pedicle length. Our complications included one return to the operating room for early flap compromise requiring reinset of the flap (one in 20 breasts). Six donor
sites have proved problematic (six of 30 legs), with either ongoing seromas, pain, or wound dehiscence. Two thighs required reoperation and two thighs (one patient) required prolonged vacuum-assisted closure. Cadaveric Study Our novel use of the branches from the gracilis pedicle to enable flap-to-flap anastomoses and therefore facilitate double flaps led us to ponder the variability of the branching patterns. As we increasingly used this technique, it was noticed that some patients had several potential branches that could be used for vascular anastomoses, whereas others had very few. If we could
1319
Plastic and Reconstructive Surgery • June 2014
Fig. 4. Double flap reconstruction following implant removal using flap-to-flap technique. (Left) Preoperatively; (right) postoperatively.
predict where we might most reliably find these branches and even be forewarned of a paucity of them, this would speed up our flap harvest and enable better planning. The vascular pedicle to the gracilis, its length and origin, and its cutaneous perforator territory have already been well described.1,12–14,17,19–21 The branches from the pedicle have been described in cadaveric and clinical articles as requiring division to raise the standard flap.17,22 However, their variability and utility for flap-to-flap anastomoses have not been investigated for breast reconstruction. One cadaveric dissection article from the Asian population has looked at the feasibility of taking double functional muscle free flaps of gracilis and adductor longus using a single pedicle and has classified the pedicles according to pattern and length.23 They did not, however, specifically investigate the size of these branches and alternative branches to other muscles (such as adductor brevis or longus), which would be as efficacious as a vascular conduit for flap-to-flap designs for breast reconstruction. Method Cadaveric soft fixed specimens were studied. The dominant gracilis pedicle was identified in an atraumatic fashion and both artery and veins were dissected back to their origin, typically from the profunda vessels. Any branches were then preserved, their destination was noted, and all vessels were measured in length
1320
and diameter using digital callipers (Maplin Electronics, South Yorkshire, United Kingdom; resolution, 0.01 mm; accuracy, ±0.02 mm). All measurements were obtained by a single investigator (R.B.) and noted simultaneously by a second investigator (J.H.). Where two veins were accompanying an artery, the largest was measured. For branches to be deemed feasible for flap-to-flap anastomoses, they had to be at least 10 mm in length, at least 10 mm away from the gracilis muscle or the origin of the pedicle, and at least 1 mm in diameter to facilitate imagined use of clamps and venous coupler device in vivo. The data were recorded in an Excel (Microsoft Corp., Redmond, Wash.) database and a classification system was proposed.
RESULTS Thirty-three thighs were dissected in 19 female and 14 male specimens; 17 were left thighs and 16 were right thighs. The mean main pedicle length was 6.83 cm (range, 4.32 to 11.95 cm), with a mean arterial diameter of 2.83 mm (range, 1.9 to 3.8 mm) and a mean venous diameter of 3.24 mm (range, 1.5 to 5 mm). All specimens had some muscular branches coming from the main pedicle, but using our cutoffs for feasibility of microvascular anastomoses, we classified the patterns as either being G0, GB, GL, or GM, or a combination where several usable branches were found to different muscles. G0 was defined as having no usable branches,
Volume 133, Number 6 • Transverse Upper Gracilis Flap GB was defined as having a usable branch to the adductor brevis muscle, GL was defined as having a usable branch to the adductor longus muscle, and GM was defined as having a usable branch to the adductor longus (Fig. 5). Using our classification, 22 thighs had suitable branches for flap-to-flap anastomosis, whereas 11 did not (classified therefore as G0). An example of a GM dissection is shown in Figure 6. Of those with suitable branching patterns, four were GB, six were GL, and eight were GM. A further four thighs had more than one suitable branch (GB/GL or GB/GL/GM) (Fig. 7). Of those with usable branches, the vessel diameters and lengths are as shown in Table 2. In those specimens where both thighs were dissected (n = 24), the symmetry between sides was examined. Seven of 12 cadavers displayed symmetry in vascular branching, whereas five did not. This encompassed the full range of branching patterns; thus, symmetry appeared no more likely for G0, GM, GB, or GL thighs. If one side was G0 in these specimens (n = 10), the other side was G0 in three, GM in two, and GB in two. This means that, although in 75 percent of patients suitable branches could be found to
Fig. 5. Schematic of branches from the main gracilis pedicle to surrounding muscles. GL, usable branch to the adductor longus; GB, usable branch to the adductor brevis; GM, usable branch to the adductor magnus.
Fig. 6. Anatomical image of gracilis pedicle with large adductor magnus branch.
allow flap-to-flap anastomoses, correspondingly, in 25 percent, potentially no good branches would be available.
DISCUSSION Compared with the more established DIEP flap for breast reconstruction, there are few series and technical tips published for free transverse upper gracilis flaps. Series published range from single figures to 154, with most numbering in the teens.1–9,24 Our series adds to the literature, but is primarily a descriptive microsurgical technique and anatomical dissection article. Similar to large series of transverse upper gracilis flaps used for breast reconstruction, we report few complications with this flap and good cosmetic results.7,9,15 Our first few patients did have problems with seromas, wound breakdown, and neuropathic pain at the donor site.24,25 These issues now occur more rarely because of, we think, resuspending the superficial fascial system to the deep fascia in the groin and periosteum, keeping drains in the thigh for up to 3 weeks (patients are educated to manage their drains at home after discharge from the hospital), avoiding harvest of the long saphenous vein with the flap,25 carefully preserving the posterior cutaneous branches of the femoral nerve, and being conservative with skin paddle dimensions. This has become more possible as we continue to improve our double flap techniques, so that flap volume and skin paddle size can be maximized without having tight donor-site closure. Excellent cosmetic results can be obtained from this donor site (Fig. 8). Our microvascular modifications have evolved with time and experience. We have had no flap failures in our series of 30 flaps, which
1321
Plastic and Reconstructive Surgery • June 2014
Fig. 7. Chart showing variation in anatomical patterns among cadaveric specimens. G0, no usable branches; GB, usable branch to the adductor brevis; GL, usable branch to the adductor longus; GM, usable branch to the adductor magnus (see text for details).
Table 2. Table of Cadaveric Measurements of Vessel Branches that Fulfilled Criteria as Being Usable for Microvascular Anastomosis Branch Length (cm)
Branch Artery Diameter (mm)
Branch Vein Diameter (mm)
Type
Mean
Range
Mean
Range
Mean
Range
GB GL GM
2.98 1.37 2.57
1.19–5.72 1.03–2.05 1.1–5.75
1.66 1.69 2.02
1.21–2.55 1.09–2.50 1.09–2.70
1.91 2.13 2.04
1.08–3.28 1.05–4.10 1.29–2.60
GB, usable branch to the adductor brevis; GL, usable branch to the adductor longus; GM, usable branch to the adductor magnus.
Fig. 8. Postoperative donor sites. (Left) Anterior and (right) posterior.
perhaps reflects a mature microsurgical practice, with approximately 250 free flaps being performed in total in our unit per year. All flaps have achieved their primary aim of breast
1322
reconstruction, without recourse to supplementary flaps or implants, in contrast to a recent series quoting reconstructive failure in onethird of free transverse upper gracilis flaps.16
Volume 133, Number 6 • Transverse Upper Gracilis Flap Nine percent of the patients in the study by Buntic et al. required further augmentation with implants.24 However, we do fairly routinely perform lipofilling and/or contralateral balancing surgery as a secondary procedure to finesse the breast mound, but this is true also of our DIEP flap reconstructions. In contrast to the approach suggested by other authors,11 when bilateral stacked transverse upper gracilis flaps are used to recreate one breast mound, we prefer to use the internal mammary axis for both flaps rather than the thoracodorsal vessels in addition, as this simplifies dissection, anastomoses, and inset. Like others, we did initially use the thoracodorsal vessels or serratus branches for second pedicle anastomoses but struggled to medialize the flap without the use of vein grafts and returned a patient to the operating room because of a hematoma following early inadvertent abduction of the patient’s arm. We therefore sought these alternative approaches using the internal mammary recipients and have found these to work well. Since adopting solely the internal mammary axis, we have returned only one patient immediately to the operating room from recovery because of a problem with inset causing kinking of the vessels. This was rectified easily and we have had no flap losses. We have not needed to use vein grafts in any of our transverse upper gracilis flaps, in contrast to other series.8 End-to-side and end-to-end anastomoses to the internal mammary vessels and use of both retrograde and anterograde arterial flow are maneuvers we have performed in our series, procedures mentioned but not described in detail in other series.25 We now favor use of flap-to-flap anastomosis with an adductor branch, a technique we have not seen published elsewhere. Perhaps because of our comfort with stacking our flaps in series in this way, a large proportion (67 percent) of our patients have had bilateral flaps for reconstructing unilateral defects, which enables a mass of tissue to be transferred to rival DIEP flaps. This avoids using a vertical scar in the medial thigh used by some authors to create volume.25,26 Other series quote smaller proportions of double flaps (15 percent8 and 21 percent11). Rather than take a very large single flap, with possible adverse donor-site sequelae such as wound dehiscence or asymmetry, we have a low threshold for using these bilateral flaps. Our anatomical study was prompted because there were no previous studies looking at the precise branching patterns of the gracilis pedicle
for us to refer to. We aimed therefore to investigate whether suitable branches would universally be found, whether there was symmetry between thighs of the same patient, where the most reliable branches would likely be situated, and as a consequence whether preoperative imaging was likely to be beneficial. In our study, the mean overall length and diameter of the main pedicle was similar to that previously published,14,23 which lends weight to the accuracy of our measurement. Certainly, our methods are standard in basic anatomical research, and inaccuracies were minimized by a single investigator measuring all of the parameters. We felt our cutoffs of a minimum of 1 cm for vessel branch length and 1 mm for vessel diameter (both artery and vein were required to fulfil these criteria) were reasonable for our microsurgical practice. Of those thighs that had suitable vessels for flap-to-flap anastomosis, eight had GB branches, had 10 GL branches, and nine had GM branches. However, the GB and GM branches were generally longer than the GL ones, and the vessel diameters were larger for the GM pattern. This is fortunate, as branches to the adductor magnus are more accessible than those to the adductor longus or brevis in vivo. The only study we could find that looked at the branches from the gracilis pedicle specifically investigated the possibility of transplanting both gracilis and adductor longus muscles with a single set of anastomoses.23 This was viewing the situation essentially in reverse and did not include lengths or diameters of the branches themselves or details of branches to other surrounding muscles. They also showed considerable variability, with 46 percent showing single branches to each muscle and 50 percent having multiple branches. In addition, they described short and ultrashort common pedicle branches very close to the profunda vessels, but this would also cause some problems with our inset. In our dissections where there was a GL pattern, the length from the profunda vessels to this branch was a mean of 2.97 cm (minimum, 1.19 cm), which is similar to that recorded by Sananpanich et al. in 2008. However, this length was, as expected, considerably longer in the GM specimens (mean, 4.87 cm) and shorter in the GB specimens (mean, 1.22 cm). Interestingly, 34 percent of the cadaveric thighs were G0 and therefore in vivo would not have been suitable for flap-to-flap anastomosis. However, 66 percent would have been, and of those paired thighs from single cadavers, having a G0 pattern on one side was not necessarily predictive of a G0
1323
Plastic and Reconstructive Surgery • June 2014 pattern contralaterally. Only one of the two thighs of course is required to have a suitable branching pattern for flap-to-flap anastomosis to be possible. Therefore, in these 12 cadavers with both thighs available for dissection, three would not have had suitable branches on either side (25 percent). Although these numbers are small, we could therefore estimate that 25 percent of our patients will not have suitable branches and 75 percent will. In the cases where no suitable branch is found on either side, double flaps are still possible, but alternative strategies need to be found. In our clinical series, we have encountered this also in 25 percent of our patient cohort where branches were sought, and have still managed to avoid use of the thoracodorsal vessels or vein grafts by using perforator vessels or end-to-side or retrograde anastomoses. However, it would be useful to know preoperatively which leg has suitable branches to allow more efficient flap harvest. In addition, to be forewarned of the lack of suitable branches being available on either thigh would allow the surgeon to ensure that any perforating vessels from the internal mammaries are preserved, or in the case of the requirement for retrograde anastomosis, the internal mammary vessels can be divided at the center of the dissected space rather than divided too far distally (Fig. 9). To this end, we have started to arrange preoperative imaging for our future transverse upper gracilis flap reconstruction patients. Eventually, we may evolve to using profunda artery perforator flaps27
in selected patients with suitable anatomy for a supine harvest, with the transverse upper gracilis flap as a good option in those with very posterior perforators on imaging.
CONCLUSIONS The transverse upper gracilis flap is now our second choice for autologous reconstruction of the breast when the abdomen is not available or insufficient. It does present the reconstructive surgeon with challenges, particularly in obtaining adequate volume and comfortable anastomosis given a short pedicle. This article outlines some of the novel microvascular approaches we have used to overcome these problems. By use of intraoperative observation of vascular anatomy and physiology, we have successfully adapted our standard techniques. We have optimized recipient pedicle length by means of costal cartilage removal and replacement, and performed nontraditional internal mammary and flap-to-flap anastomoses to allow double flaps to be used. The availability of branches from the gracilis pedicle to allow flap-to-flap anastomoses has not specifically been investigated before, and we present our cadaveric anatomical findings and a proposed classification system of these branching patterns. Knowledge of this detailed anatomy may allow more efficient flap raising and anastomotic planning; preoperative imaging may also be indicated.
Fig. 9. Decision-making algorithm for double transverse upper gracilis flap reconstruction. TUG, transverse upper gracilis flap; IM, internal mammary; IMA, internal mammary artery.
1324
Volume 133, Number 6 • Transverse Upper Gracilis Flap Judith E. Hunter, F.R.C.S.(Plast.) The Royal Marsden Hospital Fulham Road London SW3 6JJ, United Kingdom [email protected]
REFERENCES 1. Yousif NJ. The transverse gracilis musculocutaneous flap. Ann Plast Surg. 1993;31:382. 2. Wechselberger G, Schoeller T, Bauer T, et al. Surgical technique and clinical application of the transverse gracilis myocutaneous free flap. Br J Plast Surg. 2001;54:423–427. 3. Wechselberger G, Schoeller T. The transverse myocutaneous gracilis free flap: A valuable tissue source in autologous breast reconstruction. Plast Reconstr Surg. 2004;114:69–73. 4. Schoeller T, Huemer GM, Kolehmainen M, Otto-Schoeller A, Wechselberger G. A new “Siamese” flap for breast reconstruction: The combined infragluteal-transverse myocutaneous gracilis muscle flap. Plast Reconstr Surg. 2005;115:1110–1117. 5. Arnez ZM, Pogorelec D, Planinsek F, Ahcan U. Breast reconstruction by the free transverse gracilis (TUG) flap. Br J Plast Surg. 2004;57:20–26. 6. Schoeller T, Wechselberger G. Breast reconstruction by the free transverse gracilis (TUG) flap. Br J Plast Surg. 2004;57:481–482. 7. Fansa H, Schirmer S, Warnecke IC, Cervelli A, Frerichs O. The transverse myocutaneous gracilis muscle flap: A fast and reliable method for breast reconstruction. Plast Reconstr Surg. 2008;122:1326–1333. 8. Schoeller T, Huemer GM, Wechselberger G. The transverse musculocutaneous gracilis flap for breast reconstruction: Guidelines for flap and patient selection. Plast Reconstr Surg. 2008;122:29–38. 9. Vega SJ, Sandeen SN, Bossert RP, Perrone A, Ortiz L, Herrera H. Gracilis myocutaneous free flap in autologous breast reconstruction. Plast Reconstr Surg. 2009;124:1400–1409. 10. McCulley SJ, Macmillan RD, Rasheed T. Transverse upper gracilis (TUG) flap for volume replacement in breast conserving surgery for medial breast tumours in small to medium sized breasts. J Plast Reconstr Aesthet Surg. 2011;64:1056–1060. 11. Fattah A, Figus A, Mathur B, Ramakrishnan VV. The transverse myocutaneous gracilis flap: Technical refinements. J Plast Reconstr Aesthet Surg. 2010;63:305–313. 12. Yousif NJ, Matloub HS, Kolachalam R, Grunert BK, Sanger JR. The transverse gracilis musculocutaneous flap. Ann Plast Surg. 1992;29:482–490. 13. Coquerel-Beghin D, Milliez PY, Auquit-Auckbur I, Lemierre G, Duparc F. The gracilis musculocutaneous flap: Vascular supply of the muscle and skin components. Surg Radiol Anat. 2006;28:588–595.
14. Wong C, Mojallal A, Bailey SH, Trussler A, Saint-Cyr M. The extended transverse musculocutaneous gracilis flap: Vascular anatomy and clinical implications. Ann Plast Surg. 2011;67:170–177. 15. Pülzl P, Schoeller T, Kleewein K, Wechselberger G. Donorsite morbidity of the transverse musculocutaneous gracilis flap in autologous breast reconstruction: Short-term and long-term results. Plast Reconstr Surg. 2011;128:233e–242e. 16. Locke MB, Zhong T, Mureau MA, Hofer SO. Tug ‘O’ war: Challenges of transverse upper gracilis (TUG) myocutaneous free flap breast reconstruction. J Plast Reconstr Aesthet Surg. 2012;65:1041–1050. 17. Hasen KV, Gallegos ML, Dumanian GA. Extended approach to the vascular pedicle of the gracilis muscle flap: Anatomical and clinical study. Plast Reconstr Surg. 2003;111:2203–2208. 18. Koshima I, Yamamoto H, Moriguchi T, Orita Y. Extended anterior thigh flaps for repair of massive cervical defects involving pharyngoesophagus and skin: An introduction to the “mosaic” flap principle. Ann Plast Surg. 1994;32:321–327. 19. An X, Yue B, Lee JH, Lin C, Han SH. Arterial anatomy of the gracilis muscle as determined by latex injection and glycerin transparency. Clin Anat. 2012;25:231–234. 20. Lykoudis EG, Spyropoulou GA, Vlastou CC. The ana tomic basis of the gracilis perforator flap. Br J Plast Surg. 2005;58:1090–1094. 21. Whitaker IS, Karavias M, Shayan R, et al. The gracilis myocutaneous free flap: A quantitative analysis of the fasciocutaneous blood supply and implications for autologous breast reconstruction. PLoS One 2012;7:e36367. 22. Hallock GG. The gracilis (medial circumflex femoral) perforator flap: A medial groin free flap? Ann Plast Surg. 2003;51:623–626. 23. Sananpanich K, Tu YK, Pookhang S, Chalidapong P. Anatomic variance in common vascular pedicle of the gracilis and adductor longus muscles: Feasibility of double functioning free muscle transplantation with single pedicle anastomosis. J Reconstr Microsurg. 2008;24:231–238. 24. Buntic RF, Horton KM, Brooks D, Althubaiti GA. Transverse upper gracilis flap as an alternative to abdominal tissue breast reconstruction: Technique and modifications. Plast Reconstr Surg. 2011;128:607e–613e. 25. Saint-Cyr M, Wong C, Oni G, et al. Modifications to extend the transverse upper gracilis flap in breast reconstruction: Clinical series and results. Plast Reconstr Surg. 2012;129:24e–36e. 26. Kind GM, Foster RD. The longitudinal gracilis myocuta neous flap: Broadening options in breast reconstruction. Ann Plast Surg. 2008;61:513–520. 27. Allen RJ, Haddock NT, Ahn CY, Sadeghi A. Breast reconstruction with the profunda artery perforator flap. Plast Reconstr Surg. 2012;129:16e–23e.
1325
