Journal of Plastic, Reconstructive & Aesthetic Surgery (2014) 67, 526e532

Modification of the superior gluteal arteryperforator flap for reconstruction of sacralsores

Chin-Ta Lin, Shun-Cheng Chang, Shyi-Gen Chen,Yuan-Sheng Tzeng*

Division of Plastic and Reconstructive Surgery, Department of Surgery, Tri-Service General Hospital,National Defense Medical Center, Taipei, Taiwan

Received 11 October 2013; accepted 21 December 2013

KEYWORDSSuperior glutealartery perforatorflap;Sacral sore;Reconstruction

* Corresponding author. Division ofDefense Medical Center, No. 325, Sec

E-mail address: m6246kimo@yahoo

1748-6815/$-seefrontmatterª2013Brihttp://dx.doi.org/10.1016/j.bjps.2013.1

Summary Background: Despite advances in reconstruction techniques, the treatment ofsacral sores remains challenging to plastic surgeons. The superior gluteal artery perforator(SGAP) flap is reliable and preserves the entire contralateral side as a future donor site. Theipsilateral gluteal muscle is preserved, and the inferior gluteal artery flaps are viable. Howev-er, dissection of the perforator is tedious and may compromise the perforator vessels.Methods: Between April 2003 and March 2013, we performed two modified flap-harvestingtechniques: a rotational and a tunnel method, with only a short pedicle dissection to cover30 sacral defects. Patient characteristics including sex, age, cause of sacral defect, flap size,perforator number, use and postoperative complications were recorded.Results: All flaps survived except two, which developed partial flap necrosis and were finallytreated by contralateral VeY advancement flap coverage. The mean follow-up period was14.8 months (range, 3e24 months). No flap surgery-related mortality or recurrence of sacralpressure sores or infected pilonidal cysts were noted.Conclusions: Perforator-based flaps have become popular in modern reconstructive surgerybecause of low donor-site morbidity and good preservation of muscle. The advantages ofour modification procedure include shorter operative time, lesser bleeding and lesser pedicletrauma, which make the SGAP flaps an excellent choice for sacral sore coverage.ª 2013 British Association of Plastic, Reconstructive and Aesthetic Surgeons. Published byElsevier Ltd. All rights reserved.

Plastic and Reconstructive Surgery, Department of Surgery, Tri-Service General Hospital, Nationaltion 2, Cheng-Gung Road, Taipei 11490, Taiwan. Tel.: þ886 2 87927195; fax: þ886 2 87927194..com.tw (Y.-S. Tzeng).

tishAssociationofPlastic,ReconstructiveandAestheticSurgeons.PublishedbyElsevierLtd.All rightsreserved.2.031

Figure 1 Stages of the operation using the superior gluteal artery perforator (SGAP) flap. (A) Anatomical landmarks: the SGAemerges at the junction of the middle and medial thirds of a line drawn between the posterior superior iliac spine (PSIS) and thelateral border of the greater trochanter. (B1) The defect is smaller and comprises a skin strip between the defect and SGAP flap.(C1) The flap is sutured into the defect by tunnelling the pedicle, and the donor site is closed primarily. (B2) The flap is larger, andthe SGAP flap is designed to just close the defect. (C2) The flap is sutured into the defect by rotating the SGAP flap with the pedicleas the pivot point, and the donor site is closed primarily.

SGAP flap for reconstruction of sacral sores 527

Reconstructing a sacral defect caused by pressure sores orinfected pilonidal cysts is a common problem for surgeons.Many surgical methods have been used to correct sacralsores, including primary closure, skin grafting, local randomflaps and muscle flaps. Muscle and myocutaneous flaps havebeen used successfully for pressure sore coverage1 and arethe choice treatment for pressure sores because they pro-vide excellent blood supply and durable coverage.Conversely, limited shifting capacity, excessive blood lossand sacrifice of the muscle are the major drawbacks of thisprocedure.1e3

The superior gluteal artery perforator (SGAP) flap pro-vides a considerable amount of tissue with good vascularityto cover large sacral pressure sores in one stage and doesnot compromise the vascularity or innervation of the un-derlying gluteus maximus muscle.4 During SGAP flap har-vesting, meticulous dissection of the pedicle is required,which can be time consuming and tedious and requires goodtechnical skill. To shorten the learning curve and to makethe SGAP flap procedure simpler and faster to perform, wemade two modifications. First, we modified the flap designby using either a rotational or a tunnel method dependingon the size of the defect. Second, we performed a shortpedicle dissection without skeletonisation to prevent vesseltrauma and to shorten operative time. Here, we report ourexperience of the successful reconstruction of sacral pres-sure sores using this modified SGAP flap technique.

Vascular anatomy

The superior gluteal artery (SGA) arises from the internaliliac artery with a deep branch to the gluteus mediusmuscle, which then runs through the gluteus maximusmuscle and ends at the cutaneous arteries located mainly in

the superolateral gluteal region. The SGA can be marked onthe skin of the buttock at one-third of the way on a linedrawn from the posterior superior iliac spine to the top ofthe greater trochanter (Figure 1). The piriformis muscle islocated on the skin of the buttock at a line drawn betweenthe greater trochanter and a point half way to the sacrum.Perforators can be found in the area lateral to the SGA andabove the piriformis muscle.4 In 1993, Koshima describedthe anatomy of the SGA perforator.5 The length of thevessels varies from 3 to 8 cm and their diameters rangefrom 1 to 1.5 mm. These flaps can be nourished even withonly one perforator. In the cases of this study, the largestSGAP flap with one perforator was 12 � 14 cm.

Patients and methods

Between April 2003 and March 2013, 30 patients underwentsurgery for sacral sores using SGAP flaps, out of which 17were men and 13 were women, and their mean age was79.8 years (range, 22e92 years). The flap size ranged be-tween 7 � 6 cm and 12 � 14 cm. The length of the pedicledissection was determined by the arc of movement of theflap. For all operations, the length of the perforatordissection did not exceed 1 cm. All defects were grade III orIV pressure sores located over the sacral region classifiedaccording to Shea’s classification6 and infected pilonidalcysts. A total of 30 SGAP flap procedures were performed.The mean follow-up period was 14.8 months. Patient dataare given in Table 1.

Surgical procedure

The patient was placed in the prone position, theanatomical landmarks were drawn, and the SGA and its

Table 1 Characteristics of patients’ age, sex, cause of sacral defect, flap size, perforator number, utilisation, outcome, andfollow-up period.

Patient Sex/age(year)

Cause of sacral defect Flap size (cm2) Perforatornumber

Utilisation Outcome Follow-upperiod (m)

1 M/22 Infected pilonidal cyst 168 1 Rotation Good 242 M/28 Infected pilonidal cyst 108 1 Rotation Wound edge

dehiscence24

3 M/48 Oral cancer, bed ridden 42 1 Tunnel Good 124 M/55 Laryngeal cancer, bed ridden 72 1 Tunnel Good 185 M/62 ICH, bed ridden 56 1 Tunnel Good 246 F/66 ICH, bed ridden 60 2 Tunnel Good 127 M/68 Stroke, bed ridden 45 1 Tunnel Partial flap

necrosis24

8 F/72 Dementia, bed ridden 60 1 Tunnel Good 189 M/73 Stroke, bed ridden 96 1 Tunnel Good 1210 F/73 Stroke, bed ridden 108 1 Rotation Good 1211 M/75 Dementia, bed ridden 72 1 Tunnel Good 1812 M/75 Stroke, bed ridden 90 1 Rotation Good 2413 M/76 Stroke, bed ridden 165 1 Rotation Good 1214 F/77 Stroke, bed ridden 117 1 Rotation Good 1215 F/79 Stroke, bed ridden 42 1 Tunnel Wound edge

dehiscence12

16 F/80 Stroke, bed ridden 42 2 Tunnel Good 1817 M/80 Stroke, bed ridden 84 1 Tunnel Good 1218 F/81 Stroke, bed ridden 42 1 Tunnel Good 2419 F/82 Parkinson’s disease, bed ridden 70 2 Tunnel Good 1820 M/83 Parkinson’s disease, bed ridden 136 1 Rotation Good 1221 F/83 Stroke, bed ridden 56 1 Tunnel Good 1822 M/83 Stroke, bed ridden 56 2 Tunnel Good 1223 M/87 Dementia, bed ridden 42 1 Tunnel Good 2424 F/88 Dementia, bed ridden 56 1 Tunnel Good 1225 F/89 Dementia, bed ridden 63 2 Tunnel Good 1226 M/92 Stroke, bed ridden 154 2 Rotation Good 627 F/83 Leukaemia, bed ridden 80 1 Tunnel Good 628 M/50 ICH, bed ridden 88 1 Rotation Wound edge

dehiscence6

29 F/72 HIVD, bed ridden 117 1 Rotation Partial flapnecrosis

3

30 M/78 Dementia, bed ridden 120 2 Rotation Good 3

ICH: intracranial haemorrhage, HIVD: herniated intervertebral disc.

Figure 2 The algorithm for our surgical reconstructive plan.

528 C.-T. Lin et al.

perforators were identified using unidirectional Dopplerultrasound. The SGA perforators are situated mainly aroundthe junction of the middle and medial third of the linedrawn between the posterior superior iliac spine and thegreater trochanter.4 A template of the defect was drawn ona sterile, exposed radiograph, which helped ensure theaccurate size and shape of the recipient site and donortissue. The flap template was placed on the perforatormark; the skin paddle was designed with an extra 0.5 cmwidth around the margin of the template to ensure that theflap had sufficient skin to cover the defect without tension.To cover a smaller sacral sore (usually <8 cm in diameter),the SGAP flap was tunnelled beneath a skin strip betweenthe defect and the flap donor site. To cover a larger sacralsore (>8 cm in diameter), or when a bilateral conventionalVeY flap was considered, the SGAP flap was designed toclosely fit the defect. The defect was covered by a rota-tional SGAP flap (Figure 1). Figure 2 shows the algorithm forour surgical reconstructive plan.

SGAP flap for reconstruction of sacral sores 529

The incision was made superiorly and then continueddown through the skin, subcutaneous fat and fascia to themuscle. From there, the flap was detached from the muscleuntil the chosen perforator was encountered, usually in thefibrous perimysium. The vessel was slowly dissected outapproximately 1 cm in length by splitting the muscle fibresrather than by cutting them under loupe magnification, andthe vessel loop was placed around the perforator. Thefibrous septa of the perimysium were preserved during thedissection. Once the vessel was located, the inferior borderof the flap was incised, and the flap was detached from themuscle to form an island flap. This island flap was trans-posed into the defect by tunnelling the pedicle or byrotating the flap approximately 180�, as shown in Figure 1.Regardless of the method used, the length of the pedicledissection never exceeded 1 cm because of the vascularanatomy of the SGAP, which lay adjacent to the sacral re-gion. All donor sites were closed primarily. We preferred tofirst close the donor site to reduce the tension between theflap and the defect. Suction drainage was applied under theflap and in the donor area until <10 cc was collected in24 h. After the surgery, the patient’s position was changedevery 2 h. Then, the patient remained in the prone positionor on his/her side for approximately 2 weeks in the hospitaluntil the flap healed, after which they were discharged.Two weeks after discharge, the patient was instructed to liein the supine position during the day and in the prone po-sition during the night. Full supine bed rest was instructedfor approximately 4 weeks after the operation until goodwound healing was achieved.

Results

A total of 23 flaps had one SGA perforator, whereas sevenflaps had two perforators. None of the patients requiredconversion to a myocutaneous rotation flap. All flaps sur-vived, although venous congestion was common immedi-ately following surgery. A total of 19 SGAP flaps (patients 3,4, 5, 6, 7, 8, 9, 11, 15, 16, 17, 18, 19, 21, 22, 23, 24, 25 and27) were sutured into the defect by tunnelling the pedicles,and the other flaps (patients 1, 2, 10, 12, 13, 14, 20, 26, 28,29 and 30) were sutured into the defect by rotating theflap. Because of short pedicle dissection, all SGAP flapswere elevated within 1 h. Two flaps developed partial ne-crosis and were treated by contralateral VeY advancementflap coverage. Three patients developed minor complica-tions of partial dehiscence of the wound edge, which weremanaged by delayed primary closure. All donor sites wereclosed primarily and healed without any complication. Norecurrence was observed during the follow-up period(mean, 14.8 months; range, 3e24 months).

Case report

Case 9

A 73-year-old man with a history of stroke was referredwith a sacral grade IV pressure sore. After debridement, anSGAP flap measuring 8 � 12 cm with one perforator wasmoved into the defect beneath the skin strip and inset

without tension. The donor site was closed primarily. Thepatient was discharged to a nursing home uneventfully 2months later (Figure 3).

Case 20

An 83-year-old man with Parkinson’s disease, who wasbedridden, developed a sacral grade IV pressure sore. Afterdebridement, an 8.5 � 16 cm SGAP flap was designed tocover the defect. Because the defect was large, the SGAperforator was positioned adjacent to the defect. The SGAPwas taken from the zone immediately adjacent to thedefect without a tunnel, to place the SGAP flap base cen-trally on the perforators. With this design, the SGAP flaphad no marginal perforator. The defect was covered byrotating the SGAP flap with the perforator as a pivot point.The flap was sutured into the defect and the donor site wasclosed primarily (Figure 4). The patient showed no recur-rence at the 8-month follow-up examination.

Discussion

Common causes of sacral defects include pressure sores inparaplegic patients and infected pilonidal cysts in ambu-latory patients. Delayed wound coverage of a sacral defectcan cause progressive infections and wound pain. There-fore, surgical debridement and subsequent wound recon-struction are the best treatment for most patients with asacral defect.7,8 The gluteus maximus myocutaneous flap isthe most popular technique for closure of a sacral defectbecause it is reliable with a short learning curve for sur-geons. However, a disadvantage is that a functioning mus-cle has to be sacrificed, which could result in walkinginstability. Other disadvantages include a bulky appear-ance, limited flap transposition and unnecessary blood losswhen splitting the muscle.1e3 In some cases, the suture lineof the bilateral gluteus maximus myocutaneous flaps liesexactly at the maximal pressure point, which results inwound dehiscence in many patients during recovery.

In 1993, Koshima et al.5 discovered 20e25 perforatorssupplying the entire gluteal region and used gluteal perfo-rator flaps to cover sacral pressure sores. This flap is largeand safe and can be raised unilaterally with minimalbleeding, leaving the muscle intact with little donor-sitemorbidity. Verpaele et al.4 described the use of the SGAPflap considering the origin of the perforator arising from thesuperior gluteal artery that penetrates the gluteus maximusmuscle, to reconstruct a large midline sacral defect. Ac-cording to an anatomical study by Ahmadzadeh et al.,9 amean of 5 � 2 cutaneous perforators arising from the su-perior gluteal artery can be found in the gluteal area. Allthese perforators are myocutaneous perforators passingthrough the gluteus maximus muscle or the gluteus mediusmuscle. The diameters of the superior gluteal perforatorsranged from 0.6 to 1.0 mm.9 In our series, the mean size ofthe SGAP flaps was 80.1 cm2 (range, 42e168 cm2) and themaximum flap size supplied by one SGAP could reach12 � 14 cm.

Deep dissection of the perforator vessel from the muscleleaves a pedicle length of 8.5e10 cm, giving the flapimpressive mobility and allowing coverage of large defects

Figure 3 (A) Planning of an 8 cm � 12 cm SGAP flap for a 7 cm � 11 cm sacral pressure sore coverage. (B) The SGAP flap wasraised on one perforator. Note that the pedicle was still encased in its fibrous septum and the pedicle length did not exceed 1 cm.(C) The flap was tunnelled into the defect from the flap donor site, and the thickness of subcutaneous fat provided ample length totunnel the flap to the defect without further pedicle dissection. (D) The postoperative result 8 weeks after surgery.

Figure 4 (A) A large pressure sore located at the midline sacral region and planning of an 8.5 cm � 16 cm SGAP flap for sacralpressure sore coverage. Note that the flap was designed to just close the defect to avoid injury to the SGAP perforator duringharvesting. (B) The flap was raised on one perforator, which was still encased in its fibrous septum. Note that only a very shortpedicle was dissected. (C) The flap was rotated to cover the defect. (D) Stable results one week after surgery.

530 C.-T. Lin et al.

SGAP flap for reconstruction of sacral sores 531

with a unilateral flap.4 A long pedicle can be raised if alateral perforator is chosen, providing the flap with a largearc of movement, which allows undamaged tissue to beused from a distant non-traumatised zone in certain cases.However, in some very large defects, the margins of thedefect need to be included in the flap. In our series, weplanned the flap to be 1 cm longer and wider than thedefect around the predetermined perforator vessels. Insmaller defects, we included the perforator vessels in thecentral portion of the flap design, and the flap wastunnelled into the defect from the flap donor site. In largedefects, we included the margin of the defect in the flapand the flap was inset by rotating it.

In contrast to previous studies,4,7e17 we dissected onlyup to 1 cm of length of the pedicle. Dissection is timeconsuming and requires extreme care to prevent injury tothe perforator vessel; pedicle dissection is the rate-determining step of the SGAP flap for closure of a sacralsore. A flap that is slightly larger than the defect shortensthe length of the dissected pedicle and reduces operatingtime. The risk of perforator vessel injury would also bereduced. Verpaele et al.4 suggested that for SGAP flaps withmore than one perforator vessel, choosing a lateral perfo-rator would create the longest possible pedicle to raise theflap from the distant non-traumatised zone and allow thesurgeon to inset the flap with minimal torsion to thevascular pedicle.

We chose the medial perforator vessel as a pivot point torotate the SGAP flap and did not skeletonise the perforator.Although rotating the SGAP flap applies torsion to thepedicle, failure because of a kink at the pedicle was notnoted in any of the 11 rotating SGAP flaps. Studies havefailed to describe SGAP flap death from pedicletorsion.4,7e17 Raising the SGAP flap adjacent to the defectdoes not delay wound healing after adequate wounddebridement. Multiple perforator vessels restrict flapmobility, and fewer perforators allow the flap to movemore easily and farther.9 Using only one perforator maypreserve the viability of the flap. The additional perforatorvessels do not increase the viability of the flap, but provideadditional stretch between the flap and defect.11 Ligationof the marginal perforators may release tension withoutcompromising the SGAP flap.

SGAP flaps could cover the sacral sores without a longpedicle dissection because of the following: (1) because theSGA is often found at one-third of the way, down the linedrawn from the posterior superior iliac spine to the greatertrochanter, the perforator vessels will be close to the sacralsore; (2) a larger flap design ensures that the flap has suf-ficient skin to cover the defect without tension caused bytunnelling or rotating; (3) minimal undermining of thedefect and primary closure of the donor site place thedefect toward the flap and decrease the tension betweenthe flap and defect; (4) the bulk of the subcutaneous fat ofthe SGAP flap provides additional length and increases flapmobility; and (5) the far end of the sacral sore can becovered by rotating the SGAP flap up to 180�, thus elimi-nating the need for a long pedicle dissection to advance theflap.

In conclusion, sacral sore management is difficult, andflaps must be chosen carefully. The SGAP flap provides alarge, bulky and safe fasciocutaneous flap to cover sacral

pressure sores. The flap also minimises blood loss anddonor-site morbidity and preserves muscle function. Aswith other perforator flaps, pedicle dissection requires ameticulous dissection technique to prevent damage to theperforator vessels. We showed that a deep pedicle dissec-tion is unnecessary when surgery involves an accurateindicating perforator, adequate flap size design and accu-rate choice of either the tunnelling or the rotationaltechnique. Because raising the SGAP flap is no longer atechnique with a steep learning curve, we recommend thatit as a viable alternative for the management of sacral soresthat cannot be covered by primary closure or a local fas-ciocutaneous flap.

Funding

None.

Conflicts of interest

None declared.

Ethical approval

Not required.

Acknowledgement

Civilian Administration Division of Tri-Service General Hos-pital, National Defense Medical Center, Taipei, Taiwan.

References

1. Minami RT, Mills R, Pardoe R. Gluteus maximus myocutaneousflaps for repair of pressure sores. Plast Reconstr Surg 1977;60:242e9.

2. Parry SW, Mathes SJ. Bilateral gluteus maximus myocutaneousadvancement flaps: sacral coverage for ambulatory patients.Ann Plast Surg 1982;8:443e5.

3. Stevenson TR, Pollock RA, Rohrich RJ, et al. The gluteusmaximus musculocutaneous island flap: refinements in designand application. Plast Reconstr Surg 1987;79:761e8.

4. Verpaele AM, Blondeel PN, Van Landuyt K, et al. The superiorgluteal artery perforator flap: an additional tool in the treat-ment of sacral pressure sores. Br J Plast Surg 1999;52:385e91.

5. Koshima I, Moriguchi T, Soeda S, et al. The gluteal perforator-based flap for repair of sacral pressure sores. Plast ReconstrSurg 1993;91:678e83.

6. Shea JD. Pressure sores: classification and management. ClinOrthop 1975;112:89e100.

7. Kierney PC, Engrav LH, Isik FF, et al. Results of 268 pressuresores in 158 patients managed jointly by plastic surgery andrehabilitation medicine. Plast Reconstr Surg 1998;102:765e72.

8. Acarturk TO, Parsak CK, Sakman G, et al. Superior gluteal ar-tery perforator flap in the reconstruction of pilonidal sinus. JPlast Reconstr Aesthet Surg 2010;63:133e9.

9. Ahmadzadeh R, Bergeron L, Tang M, et al. The superior andinferior gluteal artery perforator flaps. Plast Reconstr Surg2007;120:1551e6.

10. Coskunfirat OK, Ozgentas HE. Gluteal perforator flaps forcoverage of pressure sores at various locations. Plast ReconstrSurg 2004;113:2012e7.

532 C.-T. Lin et al.

11. Meltem C, Esra C, Hasan F, et al. The gluteal perforator-basedflap in repair of pressure sores. Br J Plast Surg 2004;57:342e7.

12. Lee JT, Hsiao HT, Tung KY, et al. Gluteal perforator flaps forcoverage of pressure sores at various locations. Plast ReconstrSurg 2006;117:2507e8.

13. Leow M, Lim J, Lim TC. The superior gluteal artery perforatorflap for the closure of sacral sores. Singapore Med J 2004;45:37e9.

14. Cheon YW, Lee MC, Kim YS, et al. Gluteal artery perforatorflap: a viable alternative for sacral radiation ulcer andosteoradionecrosis. J Plast Reconstr Aesthet Surg 2010;63:642e7.

15. Hurbungs A, Ramkalawan H. Sacral pressure sore reconstruc-tion e the pedicled superior gluteal artery perforator flap. SAfr J Surg 2012;50:6e8.

16. Zeng A, Jia Y, Wang X, et al. The superior gluteal arteryperforator flap for lumbosacral defect repair: a unifiedapproach. J Plast Reconstr Aesthet Surg 2013;66:e56e7.

17. Moon SH, Choi JY, Lee JH, et al. Feasibility of a deep-ithelialized superior gluteal artery perforator propeller flap forvarious lumbosacral defects. Ann Plast Surg 2013 Oct 21 [Epubahead of print].