Journal of Plastic, Reconstructive & Aesthetic Surgery (2013) 66, 756e762Anatomical basis of the lateral superior gluteal arteryperforator (LSGAP) flap and role in bilateral breastreconstructionGeraldine Fade a, Fabienne Gobel a, Eric Pele b, Benoit Chaput c,Ignacio Garrido c, Vincent Pinsolle a, Philippe Pelissier a, Raphael Sinna d,*a Plastic Reconstructive and Aesthetic Surgery Department, University Hospital of Bordeaux, Bordeaux, FrancebRadiology Department, University Hospital of Bordeaux, Bordeaux, Francec Plastic Reconstructive and Aesthetic Surgery Department, University Hospital of Toulouse, Toulouse, Franced Plastic Reconstructive and Aesthetic Surgery Department, University Hospital of Amiens, U-RISE(Unit of Research and Innovation for Surgical Expertise), Amiens Medical School, 3 rue des Louvels, 80000 Amiens, France

Received 10 December 2012; accepted 20 February 2013KEYWORDSPerforator flap;Breastreconstruction;Autologous;Anatomical study;Gluteal artery;Bodylift* Corresponding author. Tel.: 33 0E-mail address: Raphaelsinna@gma

1748-6815/$ - see front matter 2013 Phttp://dx.doi.org/10.1016/j.bjps.2013.0Summary Introduction: Deep inferior epigastric perforator (DIEP) flap is one of the goldstandards in autologous breast reconstruction. When the abdominal tissue is not available,the superior gluteal artery perforator (SGAP) is often a second option with its drawback, espe-cially the donor-site deformity. Reports have highlighted that a higher and more lateral SGAPflap can be harvested to overcome several drawbacks of the classical SGAP, allowing in thesame procedure a body-contouring procedure.

In order to set the anatomical basis of this flap, we proposed to study the characteristics of areliable and easily identifiable superior and lateral perforator of the superior gluteal artery(lateral SGAP (LSGAP)) situated in the region of the lower body-lift resection allowing toperform bilateral breast reconstruction at the same time.Material and method: The anatomical study of 50 scans (or 100 buttocks) allows us to set fortha diagnostic assumption on the localisation of the perforator with respect to osseous landmarks(coccyx, iliac crest and great trochanter) which will be verified during the dissection of 10 ca-davers (or 20 buttocks) and during the 20 colour Doppler examination (or 40 buttocks).Results: In our computed tomography (CT) scan study, in 96% of cases, the perforator was si-tuated in a circle with a radius 3 cm with a 95% confidence interval and located at the junc-tion of the proximal thirdemiddle third of the distance summit of the posterior iliac crest(point B), most lateral point of the greater trochanter (point C). This assumption was verifiedby the cadaveric dissection and in vivo studies.3 22 66 83 09; fax: 33 03 22 66 87 55.il.com (R. Sinna).

ublished by Elsevier Ltd on behalf of British Association of Plastic, Reconstructive and Aesthetic Surgeons.2.017

mailto:Raphaelsinna@gmail.comhttp://dx.doi.org/10.1016/j.bjps.2013.02.017http://dx.doi.org/10.1016/j.bjps.2013.02.017http://dx.doi.org/10.1016/j.bjps.2013.02.017

Anatomical basis of the lateral superior gluteal artery perforator (LSGAP) flap 757Conclusion: Our study sets the anatomical landmarks of the LSGAP flap. This option allows theraising of an SGAP flap avoiding the main drawbacks of this flap and allows harvesting a flapwith the tissue that is often discarded during the body-lift procedure. 2013 Published by Elsevier Ltd on behalf of British Association of Plastic, Reconstructive andAesthetic Surgeons.Figure 1 Preoperative markings of a classical lower bodyliftand postoperative result. Notice the localisation of the scarover the gluteal region.The quality of breast reconstruction is now part of goodpatient management and plays an important role in thepatients quality of life after cancer. Deep inferior epigas-tric perforator (DIEP) flap reconstruction combines good-quality and reliable breast reconstruction with cosmeticallysatisfactory donor site. Unfortunately abdominal tissue isnot systematically available depending on the patientsmorphology or history. The superior gluteal artery perfo-rator (SGAP) flap is often a second-choice flap in autologousbreast reconstruction, as the main disadvantages of thisflap are its short pedicle, the firmer consistency, makingshaping of the breast more difficult, and finally the donor-site sequelae with an asymmetrical gluteal shape and a scarcrossing the buttock.

Based on an idea proposed by VanLanduyt1 we studiedthe possibility of harvesting a flap supplied by a lateralperforator of the superior gluteal artery, which wouldvascularise the cutaneous and adipose tissue that is dis-carded during the body-lift procedure, slightly higher andmore lateral to the classical SGAP flap donor site.2 Thepossibility of harvesting this flap would allow breastreconstruction, especially bilateral reconstruction, atthe same time than performing a lower body lift. Thisoption would allow decreasing the cosmetic sequelae atthe gluteal donor site (Figure 1). We therefore conductedan anatomical study in order to determine the charac-teristics of the superior and lateral perforator of the su-perior gluteal artery (lateral superior gluteal arteryperforator (LSGAP)) situated in the region of the body-liftresection.

Materials and methods

CT study

This study was based on retrospective review of 50abdominopelvic computed tomography (CT) angiogra-phies performed between October 2010 and March 2011in the radiology department of the Bordeaux UniversityHospital. Images were obtained with a Siemens Sensation16 CT (Siemens, Erlangen, Germany). There was nospecification whether the CT angiographies were taken inthe arterial or the venous phase because they were notdone for the SGAP planning but for other medicalreasons.

The LSGAP detected on CT scan was studied for eachpatient and for each side, recording the type of perforator,the pedicle length and the diameter of the vessels as theyemerged from the greater sciatic foramen.

The perforator was identified at its emergence from themuscle fascia (point P) with respect to the following bonylandmarks: coccyx (point A), summit of the posterior iliaccrest (point B) and most lateral point of the greatertrochanter (point C; Figure 2).

Point P was located by means of combinations of axialand multidimensional console reconstructions according totwo systems:

e its position with respect to triangle ABC, by measuringthe distances AB, AC and BC, and PA, PB and PC, eachestablished in a frontal plane and

e its position on orthonormal coordinates expressed incentimetres with its origin at the coccyx (point A).

The width and the height of the pelvis were recorded foreach patient in order to adjust the localisation of theperforator to the patients dimensions. All data were

Figure 3 Anatomical Landmarks. (SGA: Superior gluteal ar-tery) The perforator was located according to two landmarkssystem: e orthonormal coordinates with its origin at the coccyx(in blue) e ABS triangle (in red); with respect to the followingbony landmarks: coccyx (point A), summit of the posterior iliaccrest (point B), most lateral point of the greater trochanter(point C).

Figure 2 CT Scan of the gluteal region study. Visualisation ofa septocutaneous LSAGP and its emergence from the musclefascia (P) and localisation from the iliac crest (B) and the greattrochanter (C).

Figure 4 Anatomical dissection. Notice the important lenghtof the pedicle from the great ischiatic foramen to the P point.This dissection shows the septoctaneous course of theperforator.

758 G. Fade et al.recorded and processed by GeoGebra WebStart software(www. geogebra. org).

Anatomical study

We dissected the gluteal regions of three fresh cadavresand seven formalin-preserved cadavres, that is, 20gluteal regions. The gender and the width of the pelvisbetween the two anterior superior iliac spines wererecorded. The two landmark systems based on analysis ofCT scans were projected onto the skin in a frontal planeto eliminate curvature of the buttock (Figure 3). A fascio-adipose skin flap of the entire superior gluteal regionwas harvested from lateral to medial, and the classicalSGAP perforator and the LSGAP (measuring more than1 mm at their emergence from the muscle fascia) wereidentified.

The other perforator vessels included in the flap wereligated. The pedicle was injected with methylene blue. Themuscle or septocutaneous type of the perforator, thelength of the vessels and their diameter were recorded(Figure 4).

Ultrasound study

In a group of 20 volunteers (11 women and 9 men), twoplastic surgeons successively identified, on each side, thehypothetical site of the perforator according to the ABCsystem. For convenience, the orthonormal coordinateswere not applied, as it requires localisation of the coccyx.The ultrasound examination was performed, using a PhilipsHealthcare IU 22 ultrasound machine and a Philips 12.5 MHzlinear Doppler transducer. The radiologist checked whetherthe perforator was situated in the zone predicted by themodel.

Statistical analysis

Statistical analysis was performed using Microsoft OfficeExcel and Epi Info software (www.cdc.gov/epiinfo/).Groups were compared by means of analysis of variance(ANOVA) and ManneWhitney/Wilcoxon tests for quantita-tive variables and chi-square or Fishers exact tests forqualitative variables.

http://www.%20geogebra.%20orghttp://www.cdc.gov/epiinfo/

Figure 5 Position of the P point in the triangle ABC theGeoGebra WebStart software modelized a zone of detection ofP in a circle centred at the junction of the proximal third -middle third of the BC distance (perforator at its emergencefrom the muscle fascia (point P): coccyx (point A), summit ofthe posterior iliac crest (point B), most lateral point of thegreater trochanter (point C).

Anatomical basis of the lateral superior gluteal artery perforator (LSGAP) flap 759Students test for matched data was used to comparethe fat density in various sites of the same subject.

The limit of significance was p < 0.05.

Results

CT study

Fifty CT scans were studied (100 buttocks) in 31 men and 19women with a mean age of 65 years. The LSGAP was iden-tified in all patients. There was no significant difference inthe perforator anatomy between men and women. Thisperforator was clearly visualised in the subcutaneous tissueand systematically emerged from the septum between thegluteus maximus and gluteus medius muscles, although thecourse of the perforator was not formally demonstrated tobe exclusively septocutaneous. The mean diameter of thepedicle was 3.2 nm and the mean length of the pedicle was99.6 mm.

Localisation of the perforator (point P) in triangle ABC issummarised in Table 1.

GeoGebra WebStart software enabled us to modelise azone of detection of point P in a circle centred at thejunction of the proximal thirde middle third of the BCdistance with a mean radius of 1.6 cm (Figure 5). In 96% ofcases, the perforator was situated in a circle with a radius3 cm with a 95% confidence interval (92.2e99.8%).

The position of perforator P on orthonormal coordinatesrevealed a mean abscissa of 152 mm and a mean ordinate of91 mm (Table 1).

Weighting the coordinate according to the mean widthand height of the pelvis gave a weighted mean abscissa of216 mm and ordinate of 91 mm. Microsoft Excel softwareTable 1 CT scan study results.

Mean Stan

Age (years) 65 18Perforator diameter(mm)

3.2 0.4

Pedicle length (mm) 99.6 9.2BC distance (mm) 178 12.6AB distance (mm) 192 13.9AC distance (mm) 152 8.9GA distance (mm) 164 13.9GB distance (mm) 62 12.8GC distance (mm) 119 15.1Cercle radius (cm) 1.6 0.8Pelvis width (mm) 232 22.1Pelvis height (mm) 122 12.1Perforator abscissa (mm) 152 13Weighted abscissa (mm) 216 16.9Perforator ordinate (mm) 91 16.7Weighted ordinate (mm) 91 13.9Abdominal thickness (mm) 21 13.1Gluteal thickness (mm) 24 10.3DIEP fat density (Hs) 109 8.4SGAP fat density (Hs) 104 7.5LSGAP fat density (Hs) 103 8used to model the position of P on a graph revealed point Pin a zone with an abscissa ranging between 186 and 244 mmand an ordinate ranging between 61 and 115 mm (Figure 6).The position of the perforator on the 2-landmarking sys-tems was identical with a margin of error of

Figure 6 Comparation between the diameter of the perfo-rator and the thickness of the gluteal fat. There is no corre-lation between both.

760 G. Fade et al.The mean fat thickness and density of the differentanalysed area are summarised in Table 1. No significantdifference in fat density according to the various regionswas observed in women, as no significant difference in fatdensity was observed between the abdomen and the lowerpart of the buttock (p Z 0.51), between the abdomen andthe upper part of the buttock (p Z 0.10) or between thelower and upper parts of the buttock (p Z 0.13). No cor-relation was found between the diameter of the perforatorand the thickness of the gluteal fat.

Anatomical study

Ten dissections were performed on three fresh cadavresand seven formalin-preserved cadavres, corresponding tothree female and seven male subjects. The predictive pointof emergence of the classical SGAP perforator was deter-mined by the junction of the middle third and medial thirdof a line drawn between the greater trochanter and theposterior superior iliac spine. The hypothetical zone ofemergence of the LSGAP was then determined by using theABC landmark system. The site of emergence of theperforator was situated in a circle with a radius of 3 cmcentred on the junction of the proximal third e middlethird of the line between the greater trochanter and theposterior superior iliac spine (PSIS). The LSAGP was locatedin this predicted circle and, in every case, consisted of anexclusively septocutaneous perforator. On orthonormalcoordinates, the lateral perforator had a mean abscissa of15.3 cm and a mean ordinate of 11.2 cm, whereas theclassical perforator had a mean abscissa of 10.9 cm and amean ordinate of 10.3 cm. The mean length of the LSGAPpedicle was 10.7 cm and its mean diameter at its emer-gence from the fascia was 0.9 mm (Table 2). The cutaneousperfusion territory was always situated within the body-liftTable 2 Anatomical study results.

Dissections Mean Stan

Pelvis width (cm) 31.38 2.33LSGAP abscissa (cm) 15.35 2.28LSGAP ordinate (cm) 11.25 2.22Classical SGAP abscissa (cm) 10.89 2.22Classical SGAP ordinate (cm) 10.33 1.85LSGAP pedicle lenght (cm) 10.67 0.91LSGAP diameter (cm) 0.97 0.17resection zone with a mean area of 18 13 cm centred onthe perforator.

Ultrasound study

We therefore decided to use the first method of localisationof the perforator (triangle ABC), as it resulted in a smallerarea and did not require localisation of the coccyx, that is,a circle centred on the junction of the proximal third andmiddle third of BC with a radius of 3 cm. Localisation of theperforator by two plastic surgeons resulted in identicalcircles. In each case, the perforator was identified insidethe circle and the mean distance between the real site ofemergence of the perforator and the centre of the circlewas 1.6 cm. This perforator always had a high blood flowand coursed towards the intermuscular septum. Thevascular territory estimated by the subcutaneous branchesderived from this perforator was situated within the zone ofbody-lift resection.

Discussion

In this study, we successively determined the radiologicaland anatomical landmarks of what we called the LSGAP: anSGAP flap based on the most lateral and most superiorperforator arising from the superior gluteal artery.

The potential advantage of this flap would be to shift theSGAP flap harvest zone superiorly and laterally whileallowing simultaneous body lift, particularly in women un-dergoing bilateral breast reconstruction.

Bilateral breast reconstruction is increasingly indicated,either for bilateral cancers, unilateral cancer with contra-lateral prophylactic mastectomy or bilateral prophylacticmastectomy.

Most teams performing flap reconstructions withoutimplants prefer to perform a two-stage bilateralreconstruction.3e5 By contrast, DellaCroce6 showed thatsimultaneous reconstruction of both breasts allowed areduction of the total anaesthetic time with improvedsymmetry and a better cosmetic result.

However, the choice of flap can be difficult. The bilat-eral transverse rectus abdominis musculocutaneous flap(TRAM) flap raises the problem of donor-site morbidity. Thebilateral latissimus dorsi flap is poorly accepted, as it re-quires sacrifice of the two largest muscles of the body.7

The DIEP and SGAP appear to be more clearly indicatedfor bilateral reconstruction with less donor-site morbidityand high patient satisfaction.dard deviation Minimum Maximum

27 3412 217 156 158 149 120.8 1.5

Anatomical basis of the lateral superior gluteal artery perforator (LSGAP) flap 761The constancy of the LSGAP, although it is not alwaysseptocutaneous, supplying the zone of body-lift resection,indicates that this LSGAP flap can be reproducibly andreliably harvested from the body-lift resection zone. TheLSGAP flap could be used for good-quality one-stage bilat-eral breast reconstruction in combination with real donor-site cosmetic surgery and we would consider this to be thebest indication for the LSGAP flap.

The SGAP flap is not popular for several reasons:intramuscular dissection of the perforator is timeconsuming, the short pedicle makes anastomoses moredifficult and donor-site sequelae require subsequentreshaping operations. Moreover, the fat over the gluteusmaximus muscle is firmer due to the presence of numerousfibrous bands.

Harvesting of the LSGAP flap has been described onseveral occasions in the literature in order to overcome thedisadvantages of the SGAP8e13 but a detailed study of thelocalisation and constancy of this perforator have neverbeen performed.

This flap therefore avoids a number of disadvantagesrelated to the classical SGAP:

1. Dissection of the perforator is facilitated by its septo-cutaneous site in the great majority of cases. Thislateral situation allows the pedicle to be lengthenedand facilitates anastomoses and shaping of the flapwithout the constraints related to a short pedicle.

The SGAP harvested on the classical perforator has ashort pedicle, as, in the various studies,5,6,14e16 the meanlength of the pedicle ranged from 3 to 8 cm. Studies on theLSGAP perforator flap report a longer pedicle, as Guerra9,17

and Matar10 reported a pedicle length of 8e12 cm and Rad18

described a septocutaneous perforator with a length of11.5 cm on a cadavre specimen, 8.7 cm on the right and9.2 cm on the left on CT scan and 13 cm in vivo, whereasTuinder19 reported a pedicle length of 7.9 cm.

In our study, the mean length of the pedicle on CT scanwas 99.6 mm and the mean length on cadavre dissectionswas 10.7 cm. The length of the pedicle calculated on CT isunderestimated, as the two extremities of the vessel aremeasured without taking into account, assuming astraight-lined course, and the pedicle measured on cada-vres is shorter than that measured in vivo due to coldstorage.

Harvesting the LSGAP flap therefore provides a meangain in pedicle length of 4 cm (range 1.5e7 cm) comparedto the classical SGAP. However in this study only the arterialperforator was studied and not the veins.

2. Clinically, the consistency of the fat of the body-liftresection zone, although not demonstrated in our CTstudy, is softer than the fat of the gluteal region. Thischaracteristic makes the SGAP flap less malleable forreconstruction of the breast footprint compared toabdominal fat.10,13,17

3. Finally, the possibility of performing a body-lift proce-dure during bilateral breast reconstruction allows agreatly improved cosmetic result of the donor site withpreservation of the gluteal curve and the scar placed ina higher position that can be hidden in a G-string.The classical SGAP is usually harvested with a medi-ocaudal to laterocranial horizontal or oblique spindle-shaped paddle.8e12,20,21 One of its major drawbacks isthat it causes marked deformity of the buttock. In the studyby Babineaux,22 based on photographs and 162 question-naires on the main factors determining the choice of breastreconstruction, donor-site sequelae occupied an importantplace. In this study, the DIEP was the preferred flap in themajority of cases (46%), and the scar and the shape of thebuttock left by the IGAP were generally preferred to thoseobtained with the SGAP. These results highlight theimportance of the cosmetic result of the donor site duringreconstructive surgery.

Harvesting of a lateral oblique paddle, resulting in ahigher scar over the buttock region, has been proposed onseveral occasions in order to decrease the deformity of thebuttock.10,12,17,19,22

The characteristics of this LSGAP flap therefore appearto provide a real advantage compared to the classical SGAPallowing harvesting the same size of flap with less draw-back. However, analysis of the characteristics of thisperforator in the literature reveals a number of contro-versies. The main studies16,18,23,24 describe constant supe-rior and lateral perforators, but these vessels do not alwayshave a septocutaneous course. The various imaging mo-dalities do not allow clear visualisation of the perforatorover its entire course and therefore cannot confirmwhether this vessel is septocutaneous or musculocuta-neous, which is why we prefer the term superolateralperforator rather than septocutaneous SGAP flap.

Although many studies have tried to define localisationof the classical SGAP perforator,18,19 no published study hasdescribed a simple and precise system for localisation ofthe LSGAP perforator.

The two localisation systems used in the present studywere concordant, as simultaneous modelling of the twosystems in GeoGebra WebStart software resulted in thesame definition of point P with a margin of error of 5 mm.Variations due to the length of the curve of the buttockwere minimised by performing all measurements in thefrontal plane.

In terms of the diameter of vessels, Rad18 reported thatthe septocutaneous perforator was either absent or smallerthan the classical perforator of the SGAP in 39% of cases. Inour study, the diameter of the superolateral perforator as itemerges from the fascia is difficult to evaluate on CT scan,but the mean diameter on dissections was 0.9 mm (range0.82e1.5 mm).

Finally, in this surgery associated with a considerablerisk of accidental damage to the pedicle, the choice of theLSGAP provides two salvage options, two life boats,during bilateral breast reconstruction. The two flaps areharvested in the prone position, as during the first step ofbody lift. The patient is then placed in the supine positionfor micro-surgical suture onto internal mammary vessels.When this step has been completed successfully, theabdominal phase can be performed according to the clas-sical abdominoplasty technique. When a problem isencountered, the possibility of harvesting one or two DIEPflaps provides a relatively reassuring salvage solution withno loss of chance for the patient. However, like otherreconstruction options, this technique is obviously not

762 G. Fade et al.adapted to all patients, as the donor site must be suitableto allow body lift, that is, with a pinch test below the iliaccrest. Therefore a good indication would be any patientwho has a satisfactory gluteal donor site (normal or weightloss patients) so that after performing a body lift, the pa-tient has an aesthetic improvement of the gluteal region.This should be a leitmotiv when performing perforator flapbreast reconstruction.

Conclusion

This study confirms the constant presence of a supero-lateral perforator of the superior gluteal artery and de-scribes preoperative localisation of this perforator. In 96%of cases of this series, the perforator was situated in acircle with a radius of