Soft Tissue Sarcoma of the Hand or Foot: Conservative Surgery and Radiotherapy

Purpose. Conservative treatment in the form of limited surgery and post-operative radiotherapy is controversial in hand and foot sarcomas, both due to poor radiation tolerance of the palm and sole, and due to technical difficulties in achieving adequate margins.This paper describes the local control and survival of 41 patients with soft tissue sarcoma of the hand or foot treated with conservative surgery and radiotherapy. The acute and late toxicity of megavoltage radiotherapy to the hand and foot are described. The technical issues and details of treatment delivery are discussed. The factors influencing local control after radiotherapy are analysed. Subjects . Eighteen patients had sarcomas of the hand and 23 of the foot. All patients received post-operative radiotherapy, the majority receiving a dose of 60 Gy in 2-Gy daily fractions using a two-phase treatment. Results . The acute and late toxicity of treatment were within acceptable limits. The actuarial 5-year overall survival of the whole patient group was 67.6% and the local relapse-free survival was 44%.The local control was similar in tumours of hand and foot, and in patients treated at first presentation or relapse. Discussion. Post-operative radiotherapy to the hand or foot appears to be a well tolerated treatment resulting in long-term local control in a significant proportion of patients. The increased frequency of recurrence within the high-dose volume suggests the need for the use of higher total doses of radiotherapy.


Introduction
C onservative surgery com bined w ith radiotherapy is the recom m ended treatm ent for soft tissue sarcom a o f the extrem ity. 1± 3 H ow eve r, this a pp ro ach is controversial in sarcom a of the hand or foot, due to difficulty in achieving wide excision m argins and poor radiation tolerance of norm al tissues. Resection alone results in a high local recurrence rate of 40± 50% ; thus am putation is usually recomm ended as the treatm ent of choice. 4± 6 H owever, several recent reviews su ggest that com b in ed m od ality treatm en t w ith lim ited surgery plus radiotherapy can result in good local control with preservation of function. 7± 13 D ue to the rarity of soft tissu e sarcom a in the distal extremities, m ost reviews include only a small num ber of patients, treated with a variety of radiation techn iques. In addition m ost series include sarcom as of the w rist and ankle. T he ad dition al problem w ith the han d o r fo ot is the low rad iation tolerance of the skin of the palm and sole com pared to the re st of the lim b, and the lim itatio n of salvag e surger y for recurrence. In this paper we retrospectively review 41 patients with soft tissue sarcom a arising from hands or feet treated with conservative surgery and post-operative radiotherapy. The acute and late toxicity of m egavoltage radiotherapy, local control and functional outcom e, as well as feasib ility of salvage therapy, are analysed.

Patient characteristics
Between 1981 and 1997, 41 patients with soft tissue sarco m a arising from the han d or foot received rad iotherapy at the Royal M arsd en H ospital. All patients had non-m etastatic disease and treatm ent was intended to be curative. The m ean age was 39 years (range 13± 76 years). Twenty-one patients were m ale and 20 fem ale. Eighteen had tum ours of the hand and 23 of the foot.Thirty-one patients were irradiated following surgery at initial presentation and 10 after surgery for local recurrence. Full history and clinical examination were performed.The surgical notes and pathology reports were reviewed. Investigations included chest X-ray and tomographic scans of the chest to exclude metastases. The histological diagnosis of sarcoma was con® rm ed in all patients. The patient and tumour characteristics are sum marised in Table 1.

Treatment details
All patients underwent initial surgery, ranging from biopsy to w ide local excision, depending on the location and size of tum our. Four patients underwent biopsy only. Fourteen underwent intracapsular excision (i.e. excision with macroscopic residual tum our), 12 m arg inal excision (i.e. excision of gross tum our with pseudo capsule) and 11 w ide excision (excision of tum our with a m argin of norm al tissue, including 3 requiring ray am putation of a digit). Six patients underwent skin grafting as part of the initial surgical procedure. T he patients who underwent suboptim al surgery were operated on in other hospitals including centres abroad, and were referred post-operatively fo r rad io therap y. T he m argin s o f excision w ere assesse d from the histopathology report. N ineteen patients had clear excision m arg ins and 3 had close m arg ins. T he excision m argins were m icroscopically positive in 13 patients and 6 had gross residual tumour prior to radiotherapy. All received post-operative radiotherapy. Twenty-six received conventional dose fractionation, ranging from 50 G y in 25 fractions over 5 weeks to 60 Gy in 30 fractions over 6 weeks. Six patients received hyperfractionated radiotherapy to a dose of 72 G y in 60 fractions of 1.2 G y over 6 weeks, treating twice daily. N ine patients treated in the earlier part of the study were hypo fractionated to a dose of 39.6 G y in six fractions each of 6.6 Gy, given once a week.
T he technique of radiotherapy was determ ined by the site and extent of tumour. Twenty-six patients were treated using a large phase I volum e to a dose of 50 Gy in 25 fractions in the conventional group or 60 Gy in 50 fractions in the hyperfractionated group. This was followed by a smaller phase II volum e adding 10 G y in ® ve fra ction s or 12 G y in 10 fractions respectively during the sixth week. Fifteen patients received a single phase treatment throughout. All patients were treated with the lim b im m obilised in a cast (Figs 1 and 2).
Treatm ent po r tals w ere se lected accord in g to pre-operative com puterised tom ography (CT ) scans and surgical notes. Irradiation of the entire circum ference of the lim b was avoided, w ith care being taken to spare a corridor of skin and subcutaneous tissue. The wrist and ankle joints were excluded from the high dose volum e whenever possib le. In patients with involvem ent of the foot, the sole was spared whenever possible. A direct electron beam of energy ranging from 5 to 10 M eV was used in 7 patients; photon beam s of 5± 6 M eV or Cobalt-60 were used in the rem aining 34. Photon therapy techniques included parallel o pposed or obliqu e wedged beam s. T he surgical scar required bolus to build up the skin dose in only 3 patients.
T he overall treatm ent time ranged from 28 to 55 days (m edian 42 days). Two patients required treatm ent interruption due to an acute skin reaction. Six patients received chem otherapy as part of their initial therapy. T he 3 patients with Ewing's sarcom a and three w ith rhabdo m yo-sa rcom a received regim es containing vincristine, actinomycin D, adriamycin and ifosfam ide along with local radiotherapy. O ne patient with angiosarcom a had been treated with preradiotherapy intra-ar terial adriamycin.

Follow up evaluation
Patients were followed up every 2 m onths for the first ye ar an d at lon ger in ter vals th ereafte r. C linical exam ination and chest X-ray were perform ed at each visit. Further investigations were performed only when clinically indicated.

Functional status and treatment toxicity
The pre-radiotherapy functional status was retrospectively graded as norm al, m ild dysfu nction or severe dysfunction (Table 2). Acute and late morbidity of radiotherapy were scored using the RTO G criteria. Lim b oedema was scored using the N CIC late lim b oedem a scale. Function was also assessed at the time of last follow -up using the functional scales.

Statistical methods
Survival analysis was perform ed using the Kaplan± M eier m ethod and log-rank test. Analysis was done separately for overall survival and local recurrence. Survival was calculated from the date of surgery in all patients.

Results
T he m edian follow up was 65 m onths (range 3.5± 205 m o nths). Tw en ty-on e patien ts had a com pletely norm al pre-radiotherapy functional status; 13 had m ild dysfunction and 7 severe dysfunction. All patients co m pleted the p lan n ed co u r se o f rad iotherapy, although 2 required treatm ent interruption due to acute toxicity. T hirty-nine patients had acute skin m orbidity follow ing radiotherapy with 14 patients having grade 1 reactions, 20 grade 2, and 5 grade 3 reactions. Two of the 6 patients w ith skin grafts developed ulceration of the graft which was m anaged conservatively, and healed without further surgical intervention. L ate m orbidity, local control and survival were assesse d in 34 patients treated prior to 1995, who had follow-up in excess of 30 m onths.The late toxicity of treatment is sum m arised in Table 3 Table 4. H igh grade tum ours had higher rates of both local and nodal failure (42 and 13% respectively). The m arg ins of surgical excision did not appear to in¯uence local control or m etastatic rate, although these were assessed retrospectively from the pathological report in m ost cases. Histology of the prim ar y tum our did not in¯uence local control overall, although the subset of patients with Ewing's sarcom a or rhabdom yosarcom a had low rates of local recurrence (1/6 patients). Relapse rates were higher for patients treated for recurrence as com pared to treatm ent at initial presentation.
Kaplan± Meier estimates showed an actuarial 5-year ove rall su r v ival o f 6 7.6 % fo r th e w h ole g ro u p (Fig. 3). Patients with hand sarcom as had overall su r vival o f 73.3 % co m p are d to 63.6 % for foo t sarcom as (95% con® dence inter val, C I, 0.4± 4.8 ) (Fig. 4). Twenty

Discussion
Although large series have reported the results of conser vative surgery and radiotherapy for extrem ity sarcom as, patients w ith involvem ent of the hand or foot are few in num ber. 1,14 A high incidence of late m orbidity and functional de® cits, often necessitating am putation , has been reported follow in g irradiation. 15,16 In view of the low radiation tolerance of the skin of the palm and sole, radical resection, (often amputation), has been recom m ended as the prim ary treatm ent. 4,17 H owever, lim b conser vation therapy has gradually gained acceptance in this patient group. An im provem ent in radiation techniques and recognition of the im portance of excluding the entire joint from the high-dose volum e, plus sparing of a corridor of nor m al tissues, have resu lted in re duced late m orbidity from radiotherapy. T he routine use of CT and m agnetic resonance im ag ing (M RI) scans to de® ne tum our volum e m ore accurately and im proved im m obilisation techniques have resulted in a reduced high-dose planning target volum e and ® eld sizes in m any patients. T he use of three-dim ensional planning and conform al radiotherapy m ay aid in further reduction of acute and late norm al tissue dam age. 18 There are a few reports of good local control with preservation of lim b function using the com bined m odality approach which are sum m arised in Table 5.  T he usual recom m ended tumour dose for posto perative radio therap y o f so ft tissue sarc om a is 60 G y. 7 Tw ice daily fractionation has been proposed to im prove local control by perm itting dose escalation. 19 In our series, 6 patients received hyperfractionated radiotherapy to a total dose of 72 Gy over a period of 6 weeks. Acute toxicity was well tolerated although there was no app arent im provement in local control or survival. M ost studies demonstrate a higher rate of local failure in patients w ith positive surgical m argins. 20,21 H owever, w ide resection m arg in s are difficu lt to achieve in sarc om a o f the h an d o r foot du e to anatom ical constraints. The situation is analogous to  sarcom a arising in the head and neck. 22 In our sm all series, positive surg ical m arg ins were not asso ciated with signi® cant reduction in local control following post-operative radiotherapy and a similar conclusion has been reached by others. 10 In our series patients treated w ith post-operative radiotherapy following surgery for locally recurrent sarcom a had a higher rate of local failure com pared to those treated at initial presentation. This suggests the need for routine post-operative radiotherapy at the time of initial conservative surger y for hand and foot sarcom a. The skin of both the palm and sole tolerates radiation poorly. However, acute skin reac-tions in our patient group were acceptable, with only 2 patients treated to large volumes requiring interruption of irradiation. Radiotherapy was well tolerated in the 6 patients with skin grafts who suffered no late com plications. Grade 3 late skin toxicity in the form of m arked atrophy and telangiectasia developed in only 5.9% of the w hole series. Grade 3 late toxicity was not observed in subcutaneous tissue, bone or joints. T hese toxicity rates are com parable to those previously reported in the literature. Post-radiotherapy functional status was acceptable, with only 3 patients having m ajor dysfu nction which was at least partly due to surgery. N o patient required am putation for  intractable late com plication following radiotherapy. T here was no difference in the functional outcom e between patients with sarcom a arising in the hand com pared with those arising in the foot. H owever, another series did report worse functional outcome in foot sarcom as which might relate to weight bearing. 10 T he local control rate in our series is low at 60% , com pared to previous reports (Table 5). T his m ay be due to the high proportion of high-grade tum ours (91% ) in this patient group. In addition we had a large proportion of patients treated for locally recurrent tumour (24%) com pared to other series. Patients treated with radiotherapy at local recurrence have low local control rates com pared to those treated after the initial surgery. This suggests the need for the routine use of post-operative radiotherapy in these tum ours after initial surger y. T he m ajority of patients in our series received a total dose of 60 Gy in 2-G y fractions, com pared to higher total doses in other series (m edian 65 G y). T he proportion of recurrences within the Phase II volum e in patients with positive m argins suggests the need for dose escalation in this patient group, possib ly w ith conform ally shaped ® elds. Only 11 patients in this series underwent a standard operation in the form of w ide local excisio n p r io r to rad io thera p y, an d th is m ay h ave con tributed to the poor lo cal co ntrol.
T he m ajority of local recu rren ces were salva ged by further surger y. O nly 6 patients required am putation for local recurrence, of w hom 1 recu rred on fo u r o c ca si o n s. T h e u se o f p o s t-o p e r at ive rad iotherap y delayed am putatio n in this patien t gro up.
Distant m etastasis remains the major cause of treatm ent failure and d eath, especially in high-g rad e sarcom a. In our series, all patients who died of disease had evidence of uncontrollable distant m etastases. The use of a conservative approach avoided m utilating su rger y in these patien ts. T h e ro le of ad ju van t chem otherapy in hand and foot sarcom a requires fu rther evaluation but lack of effective ag ents is com m on to sarcom a at all sites.

Conclusion
C o n ser vative su rger y fo llow ed b y ra d io th era p y appears to be an effective treatment for soft tissue sarcom a of the hand or foot. Post-operative radiotherapy to a dose of 55± 60 Gy is well tolerated, w ith a low incidence of m oderate or severe late toxicity. The anatomical location of sarcomas in the hand or fo ot shou ld not preclud e the use of radiotherapy. Radiotherapy re sults in long-ter m local con trol in a sign i® cant propor tio n of patien ts. A m pu tation m ay be required for a sm all propo rtion of patients w ith i so l a ted m u lt ip le l o c a l r e cu r r e n c e s f o l l o w i n g radiotherapy.