Low Dose-Rate Interstitial Brachytherapy in Soft Tissue Sarcomas

Purpose. To assess the effectiveness of Ir-192 interstitial brachytherapy as an adjunct to wide local excision as a functionsaving strategy for soft tissue sarcomas. Subjects and methods. From September 1993 to April 1998, 20 consecutive patients diagnosed with soft tissue sarcomas were treated with a combination of wide local excision and interstitial brachytherapy. In 16 patients brachytherapy was done as an intraoperative procedure, while in four, the implant was performed post-operatively under local anesthesia. Eleven of the 20 patients also received external beam radiotherapy following the implant. Results. After a mean follow-up of 27 months (4–54) the local control rate for all 20 patients was 85% (17/20). In the 16 patients who had an intra-operative implant, local control was 94% (15/16). In the four patients who underwent a post-operative implant, local control was 50% (2/4). Actuarial 5-year survival was 90%. There were three cases (15%) of severe local complications. Conclusions. Wide local excision followed by low dose rate intersitital brachytherapy have yielded a 85% local control rate in 20 patients with soft tissue sarcomas. Local control rates were higher when the implants were done as an intra-operative procedure than as a post-operative one.


Introduction
O ver the last three decades several strategies have been used to avoid lim b am putation while preserving local control rates in soft tissue sarcom as of the extremities. These include w ide en bloc resection of soft parts, also known as com partmental resection, 1± 3 pre-or post-operative external radiation therapy, 4± 6 reg io n al in tr a-ar te r ia l in fu s io n 7 o r p erfu sio n chemotherapy, 8 and brachytherapy 9± 13 . Each of these treatm ent m ethods has advantages and lim itations.
T he role of local brachytherapy as an adjunct to radical surgery has been clearly established by several investigators including the pioneer experience of the group at M em orial Sloan-K ettering. 14± 16 In this pap er we describe our experience with low dose rate interstitial brachytherapy in a series of 20 patients with soft tissue tissue tum ors at various anatomical locations.

Patients and m ethods
Between S eptem b er 1993 an d M arc h 1998, 20 patients with soft tissue sarcom as were m anaged w ith wide local excision and low d ose rate interstitial brachytherapy as part of their treatment strategy. Patient characteristics are presented in Table 1.
T he selection criteria were that all patients w ith a soft tissue tumor who were found eligible for general anesthesia and a m ajor surgical procedure, and were found com plian t to u ndergo LD R brachytherapy under radiation precaution m easures were selected. All patients received an explanation of the procedures, and inform ed consent for surgery and radiotherapy was obtained.T here were eight m ales and 12 fem ales; mean age was 53 years (range, 12± 90). In 13 patients the tum or was located in the lim bs (right arm , two left arm , two; right forearm , one; right gluteal, one; left gluteal, one; right thigh, two; left thigh, two; left groin, one; left leg one), w hile in ® ve it was in the tr u nk an d in tw o in the h ead -a n d-n eck reg ion , including one patient w ith a leiomyosarcom a of the tongue. Two patients were in the pediatric age group. T here were nin e patients w ith stage I tum ors (T1a± 1b, T 2a, N 0), seven stage II (T2b, N 0) and four stage III patients. 17 T he grading system used was a two-grade system (low grade, G 1 and G2, well and m oderately differentiated; and high grade, G 3 poorly differentiated and G 4, undifferentiated) as recognized by the AJCC and as incorporated into the TN M stage grouping. 17 Following this classi® cation there were four low -grade tum ors (desm oid tum or, mixed liposarcom a, low-g rade ® brosarcoma and a chondroid syr ingom a of the subcutaneous tissue). There was one case of extraskeletal E wing's sarcom a/ PN ET of the anterior abdom inal wall. In 16 patients the brachytherapy was done as an intra-operative procedure im m ediately following wide local excision of the tum or or of the tum or bed after a m arg inal resection. T hese 16 patients were initially seen in a consultation before surgery/brachytherapy at a m u ltid isc ip lin a r y sarco m a clin ic in clu d in g surgeons (N.M ., M .E.) and radiation oncologists (E.R., A.K.). In the rem aining four patients, brachytherapy was done post-operatively in patients previously resected with inadequate marg ins. These four patients were referred to our clinic following a surgical procedure at another hospital.

Intra-operative brachytherapy procedures
Elective w ide excision and intra-operative brachytherapy were recomm ended in all 16 patients seen by us following a biopsy or m arg inal excision. Postoperative brachytherapy was done in the four patients operated elsewhere and in whom it was felt that an additional surgical procedure could not be done. In the initial nine patients of this series, and following the experience of Harrison et al. 16 and Gerbaulet et al., 20 we attempted to give all the radiotherapy dose with an im plant alone. We observed three cases of severe wound complications following the im plant (see Results) and thus we changed our treatment policy giving only a third of the radiotherapy dose with the im plant, and the additional two thirds by external beam techniques. U nder general anesthesia a wide local excision was perfo rm ed d issectin g throu gh th e healthy tissu e surrounding the tum or except in areas where it cam e into close proxim ity with bone or w ith the neurovascular bundle of the lim b. In these areas the tum or was peeled from these structures and brachytherapy catheters were placed above these high-risk regions.
Following tum or resection, gloves and instrum ents were changed. After the surgeon com pleted the tum or resection the tum or bed was m arked with m etallic clips for radiological iden ti® cation. U nder direct visualization a decision was m ade on the area to be im planted, attempting to cover the tumor bed with an additional 2 cm beyond the actual tum or confines. A series of parallel catheters were percutaneously inserted into the target area, placed app roxim ately 1 cm apart and secured in proper position using catgut sutures as required. T he entrance and exit points of the catheters were kept at not less than 1.5 cm from the surgical wound (Fig. 1). The wound was closed by approxim ation of soft tissues and skin over the catheters, care being taken to leave a drainage tube and thick viable¯aps. Plastic and m etallic brachytherapy buttons were anchored to the skin with loose stitches at the entrance and exit sites of the im plant.
Five to 7 days after the surgical procedure, a pair of orthogonal radiographs of the im plant, loaded with d u m m y m etal sou rces, w as o btained fo r sou rc e localization and com puterized treatment planning.
The im plant was then loaded w ith iridium -192 radioactive sources with an activity of 1.8 m Ci/seed (BEST Industries, Inc., Spring® eld, VA) not earlier than the ® fth post-operative day. The m edian tum or b ed im p lan t d o se w as 33 00 cG y (1 80 0± 4 90 0) delivered over 2± 5 days. T he dose was prescribed to an isodose located approxim ately 0.5 cm from the plane of the sources. This nom inal prescribed dose represents the m inim al dose within the target volum e, the d o se b etw een o r n earer th e so u rces b ein g sign i® can tly higher. 18 In 1 1 of 20 p atien ts th e in ter stitial im p la n t w a s u sed to d eliver p a r t (approxim ately one-third) of the irradiation dose. In these patients the m edian im plant dose was 2300 cGy (1800± 3000 group of patients the m edian im plant dose was 4550 cG y (4500± 4900). T he variations on the prescribed dose depended on the assessm ent of clinical variables such as: anatom ical site and proxim ity to sensitive structures, prior irradiation, geom etric quality of the im plant and potential risk for wound com plications. W hen the im plant plane was relatively close to the skin or neurovascular bundle, the lower dose (4500 cG y) was prescribed. In one patient with a hem angiopericytom a of the left groin resected with a positive m arg in , 4900 cG y w as p re sc rib ed . T his patien t developed a local recurrence 3 years follow ing the im plant. The average im plant dose rate was 48 cG y/h. E leve n o f 20 p atien ts (5 5% ) also received supplem entary external beam irradiation to a m edian dose of 3920 cG y (1620± 4500). N o patient received post-operative adjuvant chemotherapy. In the two pediatric patients, chem otherapy was given before the surgical brachytherapy procedure. A 15-year-o ld boy w ith a low -g rade ® brosarco m a o f the n eck received Cisplatin concomitantly with external irradiation before surgery/brachytherapy. A 12-year-old girl w ith a high-g rade ® brosarcom a of the left thigh received three co u rses of ifo sf am id e± etop osid e± m esn a before surger y w ithou t respon se. In fo ur patients the brachytherapy im plant was perform ed as a separate procedure following a surgical resection with positive or close marg ins. This was done under local anesthesia and using brachytherapy needles as guides for the nylon tubes in the usual m anner.

Results
After a m ean follow-up of 27 m onths (4± 54) the local control rate for all 20 patients was 85% (17/20). Actuarial 5-year sur vival was 90%. Two patients died; one of heart failure and the other of sepsis following infection of the surgical wound.
T hree patients eventually had failure at the im plant site. L ocal recurrences presented at 2, 3 and 4 years following brachytherapy. O ne patient developed lung m etastasis, being locally controlled at the prim ary tum or site.
In the 16-p atien ts w ho had an intra-o perative im plant local control was 94% (15/16); in the four patients who underwent a post-operative implant local control was 50% (2/4). Table 2 presents local control rate as a function of various treatm ent variables.

Complications
All patients developed variable degrees of erythem a of the skin overlying the im plant site 7± 10 days follow ing brachytherapy. This subsequently subsided.
Eleven patients (55% ) developed chronic radiation changes at the im plant site (NCI-C TC grade 2± 3 toxicity). T his changes consisted of dryness of the skin, subcutaneous ® brosis an d telan giectasia. In several patients, these chronic radiation changes coexisted with cosm etic defects due to the surgical resection of a variable am ount of soft tissue. Six of 11 patients with chronic radiation changes received an im plant as the sole m odality of radiotherapy, while in ® ve patients the im plant (to a lower dose) was followed by external beam irradiation.
We observed three cases of severe local com plications. An 84-year-o ld diabetic patient with a large recurrence of M FH in her right gluteal area developed a wound infection and died of septic shock 2 m onths following surger y and brachytherapy.
A 12-year-o ld girl w ith a high-grade ® brosarcom a of the right thigh w hich did not respond to induction  ch em o th erap y d evelo p ed p ro g ressive skin an d su bcutaneou s n ecrosis, w ith su bsequ ent delay in wound healing and knee contracture, in the m onths follow ing treatm ent. T his patient received 4500 cG y to the 0.5-cm plane with brachytherapy alone. Soft tissue necrosis was controlled with hyp erbaric oxygen therapy and the knee function im pairm ent required prolonged rehabilitation. O ne patient developed a local abcess which resolved with incision and drainage.

Discussion
Interstitial brachytherapy with Ir-192 as an adjuvant to radical surgery has already been show n to provide good local control rates in soft tissue sarcom as. 9± 16 The use of brachytherapy as an adjuvant to surgery is based on its theoretical and practical advantages over exte r n a l b eam ir rad iatio n . T h e b ra chyth erap y catheters are inserted in the tum or bed under direct visu alization by the surgeon an d the on cologist. Therefore, the high radiation dose is given to a target volum e w hich encom passe s the area with the greatest risk o f con tain in g resid u al m icrosco pic d isease.
Because of the rapid fall-off of the dose with distance, the surrounding tissues are relatively spared and the use of low dose rate sources provides for a greater spar ing of norm al tissues.
In a prospective random ized trial that included 164 patients, H arrison et al. 16 obtained a local control rate of 82% w ith com plete resection followed by Ir-192 brachytherapy, com pared to 69% with surger y alone. In this trial the im provem ent in local control was limited to patients with high-g rade histology, while patients with low-grade tum ors did not bene® t. Adjuvant brachytherapy improved local control, but did not have any signi® cant im pact on distant m etastases or tum or-related m ortality. T he role of radiation therapy in low -grade tum ors has not been clearly elucidated. For relatively sm all lesions that have been completely excised with negative m arg ins, postoperative irradiation is usually not recom m ended since a substantial fraction of these patients are cured by the surgical procedure. In these lesions radiotherapy is usually reserved for patients with positive m argins, deep lesions that are difficult to follow or questionable m arg ins in a location in which a local recurrence would require am putation. However, histologically low -g rade tum ors like ® brosarcoma grade I (aggressive ® brom atosis, desm oid, dermato® brosarcom a protuberans) have a typical 50% lo cal recu rren ce rate follow ing adequate su rger y alone. 19 In the present series brachytherapy was used in three patients with low-grade lesions: one with m ixoid liposarcom a, a low-grade ® brosarcom a of the neck in a 15-year-old and a case of desm oid tumor (® brosarcoma grade I) that recurred for the third time. In these three patients local control was obtained.
T he M em orial experience in the early years of their trial 16 showed m ore m ajor wound com plications (11 of 23 patients) when the im plants were loaded within the ® rst 5 post-operative days, than w hen the load in g was d one after the ® fth postoperative day (three of 21 patients). Following these results we loaded our im plants not earlier than the ® fth post-operative day.This timing allows the proliferative phase of wound healing to proceed without being impaired by radiation-induced reduction in ® broblast populations. We observed severe wound complications in two cases with probable predisposing factors. An 84-year-old patient with diabetes and ischemic heart disease was treated for a large (63 93 11 cm) recurrence of high-grade M FH in her gluteal region. This patien t rem ain ed bedr idd en after th e su rger y/ brachytherapy procedure. She subsequently developed wound infection and died of sepsis. A 12-year-old girl with a high-g rade ® brosarcom a of the thigh had previously received three courses of ifosfam ide± etoposide± m esna com bination chem otherapy without response. In this patient a wide local excision of the tum or w as followed by fu ll-dose brachytherapy (4500 cGy) to the tum or bed. The patient's age and previous chem otherapy m ay have been predisposing factors for the local com plication. In the largest published series of low dose rate brachytherapy in children G erbaulet et al. 20 treated 45 children giving prescribed doses of 6000± 7500 cGy (Paris system ). After a m ean follow -up of 5 years, 78% of the patients were alive with no evidence of d isease. A severe com plicatio n rate o f 18% w as observed in six of 33 evaluable patients. T hese six children received doses between 5800 and 7500 cGy to the pre sc riptio n iso do se , an d in these cases, sequelae can probab ly be related to the high doses delivered with the im plants. Being essentially a local m o dality tre atm en t, brachyth erap y success rates should be assesse d in term s of local control. Current resu lts are show ing bene® ts for this radiotherapy m odality in interm ediate-and high-g rade sarcom as following w id e local excision . However, it is still unclear whether brachytherapy should be used alone or in com bination with external beam techniques: and, if brachytherapy can be used alone, what would be the m inim al effective dose to m aintain current local control rates while m inim izing wound com plications? We feel that these two questions should be addressed and answered in future clinical trials. In addition, there is an ongoing debate about the im pact of local control on the developm ent of distant m etastases and disease-sp eci® c survival in patients with soft tissue sarcom as.
The im portance of obtaining negative m arg ins in the resection of soft tissue tum ors is well established. It is noteworthy that our three patients who failed locally had been resected with inadequate m arg ins (<5 m m or positive), while we noticed no local recurrences in the 15 patients w ho had pathologically adequate m arg ins (>5 m m ) ( Table 2).
We stro ngly believe that successfu l re su lts of this treatm en t approach depen d on attention to technical details su ch as: wid e resection w ith n egative m argins; m arkin g of the tum or bed w ith m etal clips; ad equ ate coverage o f the tu m or bed an d m arg in s w ith the im plant; loading of the im plant no t earlier than the ® fth po st-operative day; carefu l do sim etr y; an d d ose prescription tailored to the clinical situ ation .