Journal of Plastic, Reconstructive & Aesthetic Surgery (2013) 66, 1575e1579

Reconstruction of sarcoma defects followingpre-operative radiation: Free tissue transferis safe and reliable

W.A. Townley a,c,*, E. Mah a,c, A.C. O’Neill a,c, J.S. Wunder b,c,P.C. Ferguson b,c, T. Zhong a,c, S.O.P. Hofer a,c

aDivision of Plastic and Reconstructive Surgery, University Health Network, Toronto General Hospital,Toronto, Ontario, CanadabUniversity Musculoskeletal Oncology Unit, Mount Sinai Hospital, Toronto, Ontario, CanadacDepartment of Surgery, University of Toronto, Toronto, Canada

Received 26 March 2013; accepted 12 June 2013

KEYWORDSSarcoma;Free flapreconstruction;Pre-operativeradiation

* Corresponding author. Toronto GenE-mail address: willtownley@hotm

1748-6815/$-seefrontmatterª2013Brihttp://dx.doi.org/10.1016/j.bjps.2013.0

Summary Background: Neoadjuvant radiotherapy followed by surgical resection and soft tis-sue reconstruction provides the best possibility of achieving superior limb function in soft tis-sue sarcomas. The aim of this study was to report our experience of free flap microsurgicalreconstruction of recently irradiated soft tissue sarcoma defects.Methods: A retrospective study of microsurgical outcome in consecutively treated extrem-ity and trunk sarcoma patients undergoing free tissue transfer between 2007 and 2012 wasconducted from a prospectively collected database. Outcomes in pre-operatively irradiatedpatients were compared with non-irradiated patients. Demographic data, operative de-tails, limb salvage rate, post-operative including microsurgical complications, and long-term limb function (Toronto Extremity Salvage score, TESS; Musculoskeletal Tumour SocietyRating Scale, MSTS) were recorded and analysed for differences between the two studygroups.Results: Forty-six patients underwent 46 free flaps (pre-irradiated n Z 32, non-irradiatedn Z 14) over the study period. Microvascular complications (intra-operative revision,flap re-exploration, flap loss) were uncommon and similar between the two groups (4/32and 2/14 respectively, p > 0.05). Recipient site wound healing complications (i.e.not flap related) occurred more frequently in pre-irradiated patients (16 events) comparedwith the control group (2 events, p Z 0.03). There was no significant difference inlimb salvage rate, or TESS/MSTS functional outcome scores between the two patientgroups.

eral Hospital, 200 Elizabeth Street, Toronto, Ontario, Canada. Tel.: þ1 416 340 4800x3449.ail.com (W.A. Townley).

tishAssociationofPlastic,ReconstructiveandAestheticSurgeons.PublishedbyElsevierLtd.All rightsreserved.6.029

1576 W.A. Townley et al.

Conclusions: Free tissue transfer is safe and effective in patients undergoing surgical resec-tion and reconstruction following neoadjuvant radiotherapy.ª 2013 British Association of Plastic, Reconstructive and Aesthetic Surgeons. Published byElsevier Ltd. All rights reserved.

Introduction

Soft tissue sarcomas (STS) are a rare and diverse group ofmalignant mesenchymal tumours that account for approx-imately 1e2% of all adult cancers.1 Current standardtreatment involving wide tumour resection and pre or post-operative radiotherapy can achieve high rates of localcontrol and limb salvage with reasonable long-term sur-vival.2 Microsurgical reconstruction with free tissue transferhas an increasing role in limb preservation and achievinggood functional outcomes especially in large extremity tu-mours requiring complex resections.3,4

The timing of radiotherapy in the treatment of pa-tients with soft tissue sarcoma remains controversial.Radiation protocols typically involve either pre-operativeor post-operative external beam radiotherapy or brachy-therapy. Pre-operative irradiation carries the benefit of alower radiation dose delivered to a smaller radiationfield and the potential for downsizing of tumour priorto resection as well as better long-term functional out-comes.5 However, these positive attributes of pre-operative radiation are counterbalanced by greaterwound healing complication rates compared with post-operative radiotherapy.6 In terms of microsurgicaloutcome, there are additional concerns about the impactof radiotherapy on the quality of recipient vessels,recipient bed and ultimately flap survival. Most surgeonswould suggest that an irradiated field presents a rela-tively unfavourable environment for free tissue transferalthough limited clinical data exists in sarcoma recon-struction to support this supposition.

Our sarcoma unit delivers a multi-disciplinary teamapproach to management of patients with sarcomas thatfavours, when possible, pre-operative radiotherapy fol-lowed by surgical resection and soft tissue reconstruction.The aim of this study was to report our experience with freeflap microsurgical reconstruction of irradiated soft tissuesarcoma defects.

Materials and methods

A retrospective review was performed to study micro-surgical outcome in patients with extremity or truncalsoft tissue sarcoma. Consecutive patients undergoingfree tissue transfer by the senior author (SOPH) forsarcoma-related defects at the musculoskeletal centre atMount Sinai Hospital, Toronto between 2007 and 2012were selected for inclusion in the study. Patients wereidentified from a prospectively maintained sarcomadatabase. Both patients with STS and bone sarcomaswere selected. Patients with bone sarcomas who did notreceive radiotherapy but did require a free flap were also

included for comparative analysis, in order to increasethe size of the control group. Therefore, the controlgroup included patients with soft tissue sarcoma whoreceived post-operative irradiation following free flaphealing, as well as patients with bone and soft tissuesarcoma who did not receive any radiotherapy.7 The bulkof our practice includes extremity and trunk tumours.Head and neck sarcomas and intraabdominal sarcomaswere not included in the study. Approval for this inves-tigation was obtained from our Research Ethics ReviewBoard.

The sarcoma database, clinic reports and operativenotes were reviewed to obtain demographic and clinicalinformation. Oncologic data was collected including histo-pathological diagnosis based on the WHO classification,timing and duration of radiotherapy and need forchemotherapy.

The pre-operative radiotherapy protocol in our sarcomaunit involves administering 50 Gy of external beam radia-tion in 25 fractions over 5 weeks. If there are no significantcutaneous complications, resection and reconstruction iscarried out 4e6 weeks later. The timing of surgical inter-vention is a balance between maximising the oncologicalbenefit of radiation treatment yet allowing sufficient timeto elapse for the skin and soft tissues to recover from theinsult and thereby reduce wound-healing complicationsfollowing surgery.

Operative details were carefully documented includingtumour-related characteristics (site, size of defect), flapselection and microsurgical information (recipient vessels,anastomotic technique, need for vein graft or intra-operative revision). Outcome parameters included micro-surgical complications (return to theatre, flap failure etotal or partial), primary wound healing complications(infection, dehiscence, haematoma, delayed adjuvanttherapy) and reconstructive success (limb salvage, func-tion). The Musculoskeletal Tumour Society Rating Scale(MSTS, a clinician-rated scale scoring 0 to maximum 35) andthe Toronto Extremity Salvage Score (TESS, a patient-reported questionnaire scoring 0 to maximum100) wereused to assess limb function in patients with extremitysarcomas who had completed at least 12 months follow-up.8,9

Statistical analysis was conducted using the GraphPadonline statistical software tool (http://www.graphpad.com/quickcalcs). A two-tailed Fischer’s exact test wasselected to compare complications in pre-irradiated andnon pre-irradiated groups, whereas a two-tailed Man-neWhitney test was selected to compare demographics(age, follow-up), flap surface area and functional outcome.Data are expressed in terms of the mean � standard devi-ation (SD). A p-value of <0.05 was considered statisticallysignificant.

Table 2 Table illustrating distribution of flaps betweenpre-irradiated and control groups. ALT Z anterolateralthigh flap; LD Z latissimus dorsi.

Pre-irradiation Control

Flap selection

ALT 21 7Radial forearm 2 0LD 3 4Rectus abdominis 5 1Others 1 2

Free flap surgery in the neo-irradiated sarcoma bed 1577

Results

Over the study period, 215 patients required flap recon-struction (local, regional or free) for sarcoma-associateddefects at Mount Sinai Hospital, Toronto. Forty-six of thesepatients (21%) underwent free flap reconstruction by thesenior author thereby satisfying the inclusion criteria forthe study. Of the study group, 32 patients (age 57 � 18years, mean � SD) underwent sarcoma resection and freeflap reconstruction after neoadjuvant radiotherapy treat-ment. The control group comprised 14 patients (no radio-therapy e 13 patients, post-operative radiotherapy e 1patients) with a lower mean age at presentation (43 � 16years) reflecting the inclusion of bone sarcomas in thisgroup although the difference was not significant(p Z 0.06).

Table 1 illustrates the anatomical distribution, patho-logical diagnosis, duration of follow-up and mean defectsize of tumours in the pre-irradiated and control groups.Three patients in the pre-irradiated group also receivedpre-operative chemotherapy compared with 4 in the con-trol group (p Z 0.18).

The mean soft tissue defect surface area in the pre-irradiated group was similar to the control group. Table 2demonstrates flap selection in both study groups. Rectusabdominis flaps included those with muscle as well asmusculocutaneous (VRAM, TRAM) flaps. Similarly, bothmuscle and musculocutaneous latissimus dorsi flaps (LD)were grouped together. The ‘other’ group listed in Table 2included vastus lateralis (1 flap, pre-irradiated group), TFL(1 flap, control group) and one free fibula (control group).

In terms of microsurgical technique, vein grafts wererequired in 2 cases in the pre-irradiated group comparedwith none in the control group. The arterial anastomosisrequired intraoperative revision in 3 of 32 (9.4%) patientswho received pre-operative radiation therapy comparedwith one case (7.1%) in the control group (p Z 0.63). It isour preference to use a venous coupler for end-to-endvenous anastomoses. The mode coupler size in bothgroups was the same (2.5 mm) suggesting that the selectedrecipient veins in the irradiated bed were similar in calibre

Table 1 Table illustrating tumour site and diagnosis in thepre-irradiated and control groups.

Pre-irradiation Control p Value

Number of patients 32 14Anatomical location

Upper extremity 12 5 1.00Trunk 4 2 1.00Lower extremity 16 8 0.75

Pathological diagnosis

Soft tissuesarcoma

32 7 <0.05

Bone sarcomas 0 7 <0.05Follow-up duration(months)

52 � 21 40 � 16 0.26

Defect size (mean) 289 � 209 cm2 213 � 169 cm2 0.16

to the radiation-naive wounds beds. In the majority ofcases, a single vein was selected to drain the flap, however,two veins were deemed necessary more frequently in thepre-irradiated group (7 of 32, 21.8%) compared with thecontrol group (1 of 14, 7.1%) although the difference wasnot significant (p Z 0.40).

Table 3 compares outcome in the two study groups.There was no significant difference in flap complicationsbetween patients receiving neoadjuvant radiotherapy(total 4/32; 2 re-explorations, 1 total flap loss, 1 partial flaploss) and the non-irradiated group (total 2/14; 1 re-exploration, 1 partial flap loss). Wound healing complica-tions at the sarcoma resection site were more common inthe irradiated group (1 haematoma, 7 dehiscences, 6 earlyinfections, 1 chronic infection) compared with the non-irradiated group (1 dehiscence, 1 chronic infection).Donor site wound complications were uncommon in bothgroups.

One case of total flap loss occurred in the pre-irradiatedgroup (radial forearm to hallux) - the result of a venousthrombosis, which was not salvageable despite two earlyre-explorations. A single vein was initially selected to drainthe flap. Eventually, the wound healed with a split thick-ness skin graft. A second case in this group (rectusabdominis to pretibial region following resection of synovialsarcoma) suffered partial muscle necrosis with secondaryinfection that required multiple wound debridements.Eventually, the wound healed following a period of topicalnegative wound pressure treatment. In the control group,one flap was re-explored on the first post-operative day dueto flap congestion. The venous anasomtosis was revisedwith a vein graft and the flap subsequently healeduneventfully.

Table 3 Table demonstrating complications in the pre-irradiated and control groups.* indicates a statistically sig-nificant difference between the two study groups.

Pre-irradiation Control p Value

Complications

Flap 4 2 1.00Wound 16 2 0.03

Function

Limb salvage 28/28 12/12 1.00TESS 81.7 (n Z 15) 92.4 (n Z 6) 0.26MSTS 87 29.7 (n Z 15) 29.8 (n Z 6) 1.00

1578 W.A. Townley et al.

In the pre-irradiated group, 28 of 32 patients had ex-tremity tumours and could therefore be evaluated for limbfunction compared with 12 of 14 patients in the controlgroup (Table 3). Limb salvage was achieved in all extremitysarcoma reconstructions at follow-up. Functional outcomeas measured by the TESS and MSTS scores were similar inboth groups although only about half of patients wereeither eligible (extremity sarcomas, minimum 12 monthspost-operative) or available for formal assessment atfollow-up.

Discussion

The present study set out to establish the safety of freetissue transfer in soft tissue sarcoma patients receivingneoadjuvant radiotherapy, a critical component of themodern approach to limb-sparing sarcoma treatment. Weanalysed our results from the last 5 years, which delivered alarge volume of extremity sarcomas with free flap recon-struction compared with previous published reports. Theresults of this study suggest that free tissue transfer ishighly successful for extremity STS reconstruction in pre-operatively irradiated fields.

The control group for this study consisted predominantlyof STS patients who received either post-operative irradi-ation following wound and flap healing, or no radiation atall. However to increase the size of the control group, wealso included patients with extremity bone sarcomas whodid not receive radiation but similarly required free flapsfor soft tissue reconstruction. Resection of primary bonesarcomas typically leaves large, complex soft tissue defectssimilar to the situation for STS. Control patients were wellmatched with the pre-irradiated group for anatomical dis-tribution of tumours.

The concern of adverse microsurgical outcomes in pa-tients receiving neoadjuvant radiotherapy is genuine. Onthe anecdotal level, the irradiated wound often presents avisibly unfavourable environment to the surgeon withfibrotic, oedematous tissues and fragile recipient vesselsprone to tearing. At the cellular level, in vitro studies havedemonstrated that a course of radiotherapy induces anupregulation of cytokines in the endothelium and accumu-lation of inflammatory cells.10,11 Long-term, the changesare similar to an atherosclerotic process with intimalthickening and lipid accumulation.12 Indeed, early experi-mental models suggested that microvascular complicationsin irradiated fields were unacceptably high with flap failurerates of up to 50%.13,14

We know that macrovascular vessel reconstruction isfeasible in the irradiated sarcomatous limb althoughassociated with increased post-operative complications.15

In terms of microvascular reconstruction, evidence fromclinical studies suggests that free tissue transfer in irra-diated fields can be successful.16 Mulholland and col-leagues reported a large series of head and neck free flapreconstructions and noted an equivalent failure rate be-tween irradiated (3.5%) and non-irradiated (2.9%) flaps.17

Similar findings have also been demonstrated in breastreconstruction.18 Our own data from the current studysupports this view for extremity sarcomas as well. Chaoand colleagues reported that flap failure was independent

of radiation status in sarcoma reconstruction althoughin contrast to our study cohort, their series includeda large proportion of patients with head and neckreconstructions.19

However, other investigators have identified a strongcorrelation between pre-operative radiotherapy and sub-sequent free flap failure as well as free flap-relatedcomplications.20e22 Some studies suggested that a keystrategy to successful free tissue transfer is selectingrecipient vessels outside of the irradiated field. This is amore practical approach in head and neck reconstructionwhere there are a plethora of available recipient vessels.However, in limb reconstruction, the irradiated field oftenspans the entire limb circumference and getting out of theirradiated field may require the use of vein interpositiongrafts, which themselves can increase free flap failurerate.23 In both of our study groups, vein grafts were rarelyindicated suggesting that recipient vessels were deemedadequate in the majority of reconstructions regardless ofradiation status and despite often extensive, complex re-sections. Similarly, the comparative venous coupler datasuggested that adequate calibre recipient veins wereavailable in both irradiated and radiation-naı̈ve tissues. Itwas our preference to select flaps with long pedicles wherepossible (eg ALT, LD) to allow flexibility in planning theanastomotic sites.

Most clinical evidence related to free flap reconstructionin irradiated fields derives from head and neck recon-struction of non-sarcomatous tumours, however, all thecases in our series were either located in the extremity ortrunk. Limb reconstruction presents its own unique set ofchallenges and it is therefore not clear that the findingsfrom non-limb series can be readily extrapolated.Furthermore, although sarcomas are extremely rare ma-lignancies, our experience in a regional specialty referralcentre suggests that the majority of defects requiringresurfacing can be managed with regional flaps, rotating inrobust, well-vascularised tissue to facilitate uncomplicatedprimary wound healing. Published free flap case seriesrelated to sarcomas are therefore necessarily small and acriticism of our paper may be that it is too underpowered toyield true statistical differences. However, the fact thatfew microvascular complications occurred in a consecutiveseries of 32 flaps in irradiated sarcoma defects itself is animportant finding.

The timing of soft tissue reconstruction in relation toradiation treatment is likely to be a key factor in free flapoutcomes. In Mulholland’s series, it was noted that withinthe irradiated group, an increased time interval betweenradiotherapy and microsurgery correlated directly withflap failure.17 Conversely, Baumann and colleaguesdemonstrated that in abdominal-based free flap recon-struction of the breast, flap failure was more common ifreconstruction occurred within 12 months of radiationcompared with after 12 months.24 Microvascular recon-struction is therefore likely to be time sensitive toradiotherapy treatment although the exact associationmay also be tissue dependent. Our treatment protocolfor soft tissue sarcoma necessitating a free flap usuallyinvolves early surgical intervention following pre-operative radiation. Therefore, early reconstruction islikely to occur prior to the development of vessel changes

Free flap surgery in the neo-irradiated sarcoma bed 1579

and tissue fibrosis, which develop late in the irradiatedfield.

Although the focus of our study was on microsurgicaloutcome, we also assessed post-operative adverse compli-cations and long-term function. Wound healing complica-tions such as infection and dehiscence were relativelycommon leading to delayed discharge. This occurred muchmore frequently in the pre-operatively irradiated group.However, the finding of increased wound healing compli-cations following pre-operative radiation was not unex-pected, and has been previously reported our group andothers.6,25 Interestingly, functional outcome as describedby TESS and MSTS scores was not negatively affected bypre-operative treatment with radiotherapy, also in keepingwith the results of our previous investigations.5,26

Soft tissue sarcoma resection on the trunk and extremityresults in large complex defects, which are frequentlyamenable to direct closure or resurfacing with local orregional flaps. In select cases, free tissue transfer may berequired. Treatment with pre-operative radiotherapy is animportant component of limb-sparing surgery and improveslong-term limb function. Although pre-operative radio-therapy adversely affects wound healing, primary micro-vascular complications are not increased suggesting thatearly free tissue transfer is safe and effective in the irra-diated sarcoma field.

Conflict of interest statement

None of the authors has a financial interest in any of theproducts, devices, or drugs mentioned in this manuscript.

References

1. Clark M a, Fisher C, Judson I, Thomas JM. Soft-tissue sarcomasin adults. N Engl J Med 2005 Aug 18;353(7):701e11.

2. Sampath S, Schultheiss TE, Hitchcock YJ, Randall RL,Shrieve DC, Wong JYC. Preoperative versus postoperativeradiotherapy in soft-tissue sarcoma: multi-institutional anal-ysis of 821 patients. Int J Radiat Oncol Biol Phys 2011 Oct 1;81(2):498e505.

3. Davidge KM, Wunder J, Tomlinson G, Wong R, Lipa J, Davis AM.Function and health status outcomes following soft tissuereconstruction for limb preservation in extremity soft tissuesarcoma. Ann Surg Oncol 2010 Apr;17(4):1052e62.

4. Grinsell D, Di Bella C, Choong PFM. Functional reconstructionof sarcoma defects utilising innervated free flaps. Sarcoma2012 Jan;2012:315190.

5. Davis AM, O’Sullivan B, Turcotte R, et al. Late radiationmorbidity following randomization to preoperative versuspostoperative radiotherapy in extremity soft tissue sarcoma.Radiother Oncol 2005 Apr;75(1):48e53.

6. O’Sullivan B, Davis AM, Turcotte R, et al. Preoperative versuspostoperative radiotherapy in soft-tissue sarcoma of the limbs:a randomised trial. Lancet 2002 Jun 29;359(9325):2235e41.

7. Hogendoorn PCW, Athanasou N, Bielack S, et al. Bone sar-comas: ESMO Clinical Practice Guidelines for diagnosis, treat-ment and follow-up. Ann Oncol 2010 May;21(Suppl. 5):v204e13.

8. Enneking W. Limb salvage in musculoskeletal oncology.New York: Churchill Livingston; 1987. p. 626e39.

9. Davis AM, Wright JG, Williams JI, Bombardier C, Griffin A,Bell RS. Development of a measure of physical function forpatients with bone and soft tissue sarcoma. Qual Life Res 1996Oct;5(5):508e16.

10. Russell NS, Hoving S, Heeneman S, et al. Novel insights intopathological changes in muscular arteries of radiotherapy pa-tients. Radiother Oncol 2009 Sep;92(3):477e83.

11. Weintraub NL, Jones WK, Manka D. Understanding radiation-induced vascular disease. J Am Coll Cardiol 2010 Mar 23;55(12):1237e9.

12. Schultz-Hector S, Trott K-R. Radiation-induced cardiovasculardiseases: is the epidemiologic evidence compatible with theradiobiologic data? Int J Radiat Oncol Biol Phys 2007 Jan 1;67(1):10e8.

13. Tan E, O’Brien B, Brennen M. Free flap transfer in rabbits usingirradiated recipient vessels. Br J Plast Surg 1978;31:121e3.

14. Krag C, De Rose G, Lyczakowski T, Freeman CR, Shapiro SH.Free flaps and irradiated recipient vessels: an experimentalstudy in rabbits. Br J Plast Surg 1982 Jul;35(3):328e36.

15. Ghert MA, Davis AM, Griffin AM, et al. The surgical and func-tional outcome of limb-salvage surgery with vascular recon-struction for soft tissue sarcoma of the extremity. Ann SurgOncol 2005 Dec;12(12):1102e10.

16. Kroll SS, Robb GL, Reece GP, et al. Does prior irradiation in-crease the risk of total or partial free-flap loss? J ReconstrMicrosurg 1998 May;14(4):263e8.

17. Mulholland S, Boyd B, McCabe S, et al. Recipient vessels inhead and neck microsurgery: radiation effect and vessel ac-cess. Plast Reconstr Surg 1993;92(4):628e32.

18. Fosnot J, Fischer JP, Smartt JM, et al. Does previous chest wallirradiation increase vascular complications in free autologousbreast reconstruction? Plast Reconstr Surg 2011 Feb;127(2):496e504.

19. Chao AH, Chang DW, Shuaib SW, Hanasono MM. The effect ofneoadjuvant versus adjuvant irradiation on microvascular freeflap reconstruction in sarcoma patients. Plast Reconstr Surg2012 Mar;129(3):675e82.

20. Khouri RK, Cooley BC, Kunselman AR, et al. A prospective studyof microvascular free-flap surgery and outcome. Plast ReconstrSurg 1998 Sep;102(3):711e21.

21. Halle M, Bodin I, Tornvall P, Wickman M, Farnebo F,Arnander C. Timing of radiotherapy in head and neck freeflap reconstruction-a study of postoperative complications.J Plastic, Reconstr Aesthet Surg 2009 Jul;62(7):889e95.

22. Lee S, Thiele C. Factors associated with free flap complicationsafter head and neck reconstruction and the molecular basis offibrotic tissue rearrangement in preirradiated soft tissue.J Oral Maxillofac Surg 2010;68(9):2169e78.

23. Bozikov K, Arnez ZM. Factors predicting free flap complicationsin head and neck reconstruction. J Plastic, Reconstr AesthetSurg 2006;59(7):737e42.

24. Baumann DP, Crosby M, Selber JC, et al. Optimal timing ofdelayed free lower abdominal flap breast reconstruction afterpostmastectomy radiation therapy. Plast Reconstr Surg 2011Mar;127(3):1100e6.

25. Cheng EY, Dusenbery KE, Winters MR, Thompson RC. Soft tissuesarcomas: preoperative versus postoperative radiotherapy.J Surg Oncol 1996 Feb;61(2):90e9.

26. Davis a M. Function and health status outcomes in a random-ized trial comparing preoperative and postoperative radio-therapy in extremity soft tissue sarcoma. J Clin Oncol 2002 Nov15;20(22):4472e7.