Importance of Local Treatment in Pediatric Soft Tissue Sarcomas WithMicroscopic Residual After Primary Surgery: Results of the Italian Cooperative

Study RMS-88

Giovanni Cecchetto, MD,1* Modesto Carli, MD,2 Guido Sotti, MD,3

Gianni Bisogno, MD,2 Patrizia Dall’Igna, MD,1 Camillo Boglino, MD,4

Claudio Granata, MD,5 Luca Antoniello, MD,1 and Maurizio Guglielmi, MD1

Background. The goal of primary excisionin soft tissue sarcomas is the complete removalof the tumor by a nonmutilating procedure.However, microscopic residuals may be left af-ter a conservative procedure because of inad-equate preoperative assessment or difficultiesduring the operation. The purpose of this reportis to describe the treatment and the outcome inpatients, enrolled in the Italian CooperativeStudy RMS-88, with microscopic residuals afterprimary excision (IRS Group IIa). Procedure.Microscopic residuals were evident at histologyin 52 of 90 patients who had a macroscopiccomplete primary excision: 25 rhabdomyosar-comas (RMS) and 27 nonrhabdo-soft tissue sar-comas (NRSTS). Eighteen patients were treatedwith primary reexcision (PRE) and chemo-therapy (CT) using VA or IVA regimens; 27 pa-tients received radiation therapy (RT; 40 Gy)and IVA; 7 children in whom PRE was not fea-sible and RT could not be administered for

age <3 years were treated with CT (IVA) alone.Results. Of the 18 patients who underwent asuccessful PRE + CT, the local relapses were 3(16.6%); of 27 cases who had RT + CT therewere 4 local relapses (14.8%); 3 local relapsesoccurred in those 7 patients in whom CT alonewas administered (43%). Conclusions. Micro-scopic residuals after primary surgery were dif-ficult to manage because of the absence of ameasurable target. PRE represented the treat-ment of choice for children <3 years of age whocannot receive RT and for paratesticular sites.PRE and RT showed similar results in achievinglocal control in extremity and trunk sites, butthey could not always avoid local recurrence.In particular PRE was not effective in tumorslarger than 5 cm. If microscopic residuals couldnot be avoided and PRE was not possible, ad-equate RT was effective both for RMS and forNRSTS. Med. Pediatr. Oncol. 34:97–101,2000. © 2000 Wiley-Liss, Inc.

Key words: pediatric soft tissue sarcoma; surgery; radiation therapy; chemotherapy;primary reexcision

INTRODUCTION

Surgical management of soft tissue sarcomas (STS) isgenerally less aggressive in children than in adults be-cause of the chemosensitivity of the majority of thesetumors in childhood and adolescence. Local treatment,however, is one of the main factors in the outcome ofSTS, and every effort should be made to achieve localcontrol [1].

Most therapeutic protocols plan the initial resection ofthe tumor only when a complete microscopic not muti-lating excision is feasible [2,3]. Otherwise, initial surgerymust be limited to a diagnostic biopsy, and the childshould undergo a delayed excision after chemotherapy(CT) with or without radiation therapy (RT). Indeed, re-section with macroscopic residual does not offer any ad-vantage over a biopsy. Microscopic residual may be leftafter a primary excision because of difficulties during theoperation or because the sarcoma is misinterpreted as abenign lesion and resected by a shell-out procedure. Inthese cases the contribution made by CT or RT to local

control is difficult to determine because there is not ameasurable target to assess efficacy.

1Department of Pediatrics, Division of Pediatric Surgery, University ofPadua, Padua, Italy2Department of Pediatrics, Division of Onco-Hematology, Universityof Padua, Padua, Italy3Division of Radiation Therapy, General Hospital of Padua, Padua,Italy4Division of Pediatric Surgery, Pediatric Hospital ‘Bambin Gesu`,’Rome, Italy5Division of Pediatric Surgery, Pediatric Hospital ‘G.Gaslini,’Genova, Italy

Presented at the Third International Congress on Soft Tissue Sarcomain Children and Adolescents, Stuttgart, 30 April to 3 May, 1997.

Grant sponsor: Italian National Research Council; Grant number:96.00658-P.F.ACRO 39; Grant sponsor: MURST (ex. 60%).

*Correspondence to: Giovanni Cecchetto, MD, Department of Pedi-atrics, Division of Pediatric Surgery, University of Padua, Via Gius-tiniani 3, 35128 Padua, Italy. E-mail: chpedpd@child.pedi.unipd

Received 28 August 1998; Accepted 7 October 1999

Medical and Pediatric Oncology 34:97–101 (2000)

© 2000 Wiley-Liss, Inc.

We analyzed patients affected by localized rhabdo-myosarcomas (RMS) and non-RMS soft tissue sarcomas(NRSTS), enrolled in the Italian Cooperative Study (ICS)RMS-88, who had microscopic residual after primaryexcision. The results obtained with different treatmentsare discussed.

MATERIALS AND METHODS

Between January 1988 and December 1995, 298 pa-tients with a localized STS were registered in the ICSRMS-88: 201 RMS, 97 NRSTS. All these patients weregrouped according to the Intergroup RhabdomyosarcomaStaging (IRS) system as follows: Group I, complete mi-croscopic excision; Group II, excision with microscopicresidual (IIa) or with regional lymphnodes involvement(IIb) or both (IIc); Group III, initial biopsy or excisionwith macroscopical residuals.

Ninety of two hundred ninety-eight patients (53 RMS,37 NRSTS) underwent macroscopic primary excision.Fifty-two of these ninety cases (25 RMS, 27 NRSTS) hadmicroscopic residual (either found or suspected) afterthis procedure, without regional lymphnodes involve-ment (Group IIa). The NRSTS were 6 sinovial sarcomas,5 malignant peripheral nerve sheath tumors, 4 peripheralneuroectodermal tumors, 2 extraosseus Ewing sarcomas,10 other types. Microscopic residuals were defined his-tologically as neoplastic foci found on the biopsies of thetumor bed or on the margins of the specimen. Further-more, when no sufficient histological data on the com-pleteness of the operation were available, the excisionwas considered with microscopic residual.

All 52 patients with microscopic residual, 31 malesand 21 females (median age 87 months, range 12–204months), were evaluable for treatment. The pathologypanel of the study reviewed all cases.

The surgical guidelines of the ICS RMS-88 for pa-tients with microscopic residual recommended a primaryreexcision (PRE), if feasible. The primary reexcision is anonmutilating wider local excision performed prior toany other treatment with the aim of achieving the micro-scopic radicality and avoiding local irradiation. PRE wasconsidered effective if residuals were not found or werecompletely resected. Patients restaged as Group I after asuccessful PRE received only CT with vincristine andactinomycin D (VA) if they had embryonal RMS or if-osfamide, vincristine and actinomycin D (IVA) if theyhad alveolar RMS and NRSTS (Table I).

IVA regimen was utilized for all patients, when PREwas not feasible or had failed to achieve the completeexcision (Table I); moreover, an early RT (between 7 and9 weeks) was delivered to children over 3 years old,using a hyperfractionated and accelerated regimen. Twodaily fractions of 160 cGy were given up to a total dose

of 40 Gy in 2.5 weeks. RT was not recommended inpatients younger than age of 3 years, to avoid late se-quelae.

RESULTS

PRE was done in 21 of 52 patients: The interval be-tween primary surgery and PRE was 7–56 days (median29 days). Residuals were not detected in 10 of 21 (48%)patients; they were found and completely resected in 8 of21 (38%), whereas they were not completely excised in 3of 21 (14%) cases (1 thoracic wall, 1 bladder, and 1 face).The 18 STS (7 RMS, 11 NRSTS) in which the completemicroscopic excision was successful were localized inparatesticular (7), extremity (6), trunk (4), and head-neck(1) regions (Table II). Four patients were younger than 3years. The initial tumor size was >5 cm in 6 cases and <5cm in 12 (Table III). In paratesticular site, PRE wasperformed because at primary surgery the tumor masswas resected without orchiectomy or for an inadequatesurgical approach through a scrotal incision; hemiscro-tectomy was needed in 5 of 7 patients. All these 18patients received only CT according to the protocol.

The 31 patients in whom PRE was not feasible fortumor site and those 3 in whom it failed were treated withIVA regimen. Twenty-seven of these thirty-four casesreceived RT, whereas 7 of 34 did not because of age. Theinitial tumor size in patients who received RT was >5 cmin 8 cases and <5 cm in 19. In the 7 children treated withCT alone, the tumor size was >5 cm in 4 and <5 cm in 3(Table III).

At present 39 of 52 patients are in first complete re-mission (CR) with a median follow-up of 63 months(range 30–121 months). The estimated disease-free sur-vival rates were calculated with Kaplan-Meier method:

TABLE Ia. ICS-RMS-88: Chemotherapy Regimen in Group IEmbryonal RMS (×4 Cycles)*

Week

1 2 3 4

AMD × ×VCR × × × ×

*AMD, actinomycin D 1.8 mg/m2 (max. 2.4 mg); VCR, vincristine 1.5mg/m2 (max. 2 mg).

TABLE Ib. ICS-RMS-88: Chemotherapy Regimen in Group IAlveolar RMS, NRSTS, and All Group II STS (×3 Cycles)*

Week

1 2 3 4 5 6 7

IFO ×× ×× ××AMD × × ×VCR × × × × ×

*IFO, ifosfamide 3 g/m2/day × 2 days; AMD, actinomycin D 1.5mg/m2 (max. 2 mg); VCR, vincristine 1.5 mg/m2 (max. 2 mg).

98 Cecchetto et al.

At 60 months from diagnosis they are 70.7% (95% CI49–92.6) after PRE + CT, 83% (95% CI 68–98) after RT+ CT, 57% (95% CI 20.5–93) after CT alone. The 13cases with a relapse are summarized in Table IV.

The relapses after PRE were found locally in 3 of 18cases (16.6%), 2 on the chest wall and 1 on the leg, at 4,10, 18 months from diagnosis; 2 patients with primarytumor on paratesticular and extremity sites had pulmo-nary metastases at 12 and 36 months. All 3 patients withlocal relapse (LR) had a mass >5 cm in size (Table III);at PRE no residuals had been found in 2 of 3 cases,whereas residuals had been completely resected in 1 of 3.All the 4 patients aged under 3 years, who underwentPRE are in CR (1 in second CR).

Among the 27 patients who received RT + CT, 4 hadLR (14.8%) and 1 pulmonary metastases (primary onextremity) at 9, 36, 45, 60, and 27 months from diagno-sis, respectively. The patients with LR had tumors onextremities (2), trunk (1), and HNnPM (1), and 2 of themhad a tumor mass >5 cm in size.

The failures in the 7 patients, who received CT alonebecause of age, were three (43%); they were caused byan LR (1 HN-PM, 2 intrathoracic), which appeared at 7,16, and 30 months from diagnosis. All 3 patients had anNRSTS, and only 1 case had a tumor mass >5 cm.

The LR rate after PRE and that after RT were quitesimilar among RMS and NRSTS: 1 of 7 RMS and 2 of 11NRSTS after PRE, 2 of 12 RMS and 2 of 13 NRSTS afterRT. Among patients who did not receive local treatment,

being treated only with CT, those 3 with NRSTS had alocal relapse, confirming that NRSTS are less chemosen-sitive than RMS (Table IV).

All 3 patients with distant metastases died of progres-sive disease, and 5 of 10 patients with LR died of thedisease (1 after PRE + CT, 1 after RT + CT, and all 3after CT alone); the other 5 cases (2 after PRE + CT and3 after RT + CT) obtained the second CR by furthermultimodal nonmutilating therapy and are alive withoutdisease at 25–61 months from LR.

DISCUSSION

In recent years, conservative surgery for children withSTS has been more widely utilized. This increases therisk of microscopic residual disease after primary exci-sion, a difficult reality to face. The surgical oncologistshould plan a radical procedure only when it is possibleto provide a margin of normal tissue, conventionally 1–2cm, or a fascial plan surrounding all the circumference ofthe tumor. Otherwise, he should plan only an incisionalbiopsy and reevaluate the possibility of complete exci-sion after CT. The decision to renounce primary excisionis quite easy for some primary sites (orbit, paramenin-geal, intracavitary, and GU bladder/prostate) or for somehuge intracavitary tumors. It may, however, be difficultfor extremity, trunk, or nonparameningeal-head/neck sar-comas. For these locations the surgeon must take intoconsideration the size and the invasiveness of the tumorto evaluate the feasibility of a radical excision.

It may happen that a planned wide resection turns outto be a marginal excision (according to Ennekin criteria),i.e., with microscopic residual [4]. Unexpected technicaldifficulties during the operation, resulting from tumorinfiltration or extension, can be the cause of a resectionwith involved margins. Resection of a lesion thought tobe benign can also leave a microscopic residual; in fact,many STS are apparently circumscribed, and the pres-ence of an easy plane of cleavage between the tumor andthe surrounding tissue makes it easy to perform an op-eration without observing surgical oncology criteria [5].

The evidence of neoplastic foci on the margins of thespecimen or of microscopic residuals on the biopsies ofthe tumor bed considerably increases the risk of localrecurrence. In these cases the first therapeutic option isPRE, widely adopted for adults STS [6,7]. If a total ex-cision free of microscopic residual can be achieved bythis procedure, RT can be avoided.

According to the Hays et al. [8] guidelines, PRE isgenerally feasible for some locations, particularly ex-tremities and trunk, whereas it is more difficult for in-tracavitary and head/neck sarcomas. In our experience,successful PRE was performed in 6 of 15 sarcomas of theextremities and in 4 of 8 of the trunk; it was not feasible

TABLE II. Treatment According to Tumor Site*

Site Pre + CT RT + CT CT Total

Extremities 6 8 1 15Others 4 4 (1) 5 13HNnPM 1 9 (1) 0 10GUnBP 7 2 0 9Orbit 0 3 0 3HNPM 0 0 1 1GUBP 0 1 (1) 0 1Total 18 27 (3) 7 52

*HNPM, head/neck parameningeal; HNnPM, head/neck nonpara-meningeal; GUBP, genitourinary bladder prostate; GUnBP, genitouri-nary nonbladder prostate; parentheses show cases in which PRE wasinitially attempted and failed in obtaining radicality, so the patientswere treated locally by RT.

TABLE III. Treatment According to Tumor Status*

T1a T1b T2a T2b

Pre + CT 10 4 (2) 2 2 (1)RT + CT 19 (2) 8 (2) — —CT 2 (1) 1 1 (1) 3 (1)Total 31 (3) 13 (4) 3 (1) 5 (2)

*T1a, tumor confined to tissue (or organ) of origin, <5 cm in size; T1b,tumor confined to tissue (or organ) of origin, >5 cm in size; T2a, tumorinvolving contiguous structures, <5 cm in size; T2b, tumor involvingcontiguous structures, >5 cm in size; parentheses show local relapses.

Microresidual Pediatric Soft Tissue Sarcoma 99

in 12 patients and did not achieve radicality in only 1patient among those with extremity and trunk localiza-tions (Table II). Another 2 patients in whom PRE failedto achieve radicality were a child younger than the age of3 years with an RMS of the bladder for whom PRE wasconsidered the best local therapeutic option and a 14-year-old boy with a RMS of the nasolabial fold. Thelatter boy probably should have undergone RT initiallybecause of the predictable surgical difficulties in this site.All 3 of these patients are alive without disease after RT,CT, and delayed nonmutilating surgery in the first patientand after RT + CT in the other 2.

PRE was also effective in our group for those parates-ticular sarcomas in which the first surgical approach hadbeen through a scrotal incision or after a simple tumor-ectomy through a correct inguinal approach; local controlwas achieved in 7 patients by this procedure. Among the18 patients who underwent a successful PRE, 3 (16.6%)relapsed locally (in 2 of them no residuals had beenfound at PRE). All 3 cases had initially huge tumormasses; PRE resulted in radicality in all 12 cases withtumor size <5 cm, whereas it failed in 3 of 6 with tumorsize >5 cm (Table III). Probably not only the primary sitebut also the initial tumor size must be considered inplanning PRE.

Postoperative RT is commonly recommended in caseswith microscopic residuals, in children as in adults [9].Several studies demonstrated that radiation doses of 40Gy delivered to an adequate volume allowed good localcontrol in IRS Group II pediatric patients [10]. Currentlythe recommended volume is based on the initial tumorsize as delineated by CT scan or MRI before surgery,including a 2–5 cm margin. However, in the absence ofa target, it may be very difficult to establish the volumeof irradiation if accurate imaging studies have not beenperformed before surgical excision.

In an attempt to improve the therapeutic effectivenessof RT, maintaining a safety margin for late effects, thedivision of the daily dose into two smaller fractions (hy-perfractionation) has been a subject of interest in child-hood STS [11]. This regimen was adopted in the presentstudy. The main advantages of this hyperfractionated,accelerated RT would be a reduction of late effects and ashortening of the overall treatment, allowing for betterassociation with the chemotherapeutic regimen.

In our group, 4 of 27 patients (2 with a RMS and 2with a NRSTS) had an LR, which was within the RTfield in all cases. The local failure rate of 14.8% (4/27)after RT is quite similar to the rate registered after PRE.Among those 7 patients who received only CT, the localrelapses were 3. All these 3 patients, aged less than 3years, had an NRSTS. It is possible that a better initialassessment in these patients would have prevented anexcision with microscopic residual.

The cure of microscopic residual by CT alone in SIOPexperience is possible in a selected group of RMS, re-serving local treatment for relapsed cases [12]; the Ger-man CWS Group describes a high local relapse rate inSTS treated without local therapy, especially in NRSTSand alveolar RMS [13]. Our results demonstrate that mi-croscopic residual disease after primary surgery is diffi-cult to treat. Among the patients treated by PRE or RT,the 3 who had distant metastases and 2 of the 7 whorelapsed locally died of the disease. The other 5 patientsachieved a second CR with further therapy. Among the 7patients younger than 3 years, who did not receive localtreatment, 4 affected by RMS (3 embrional and 1 alveo-lar) are alive without disease and 3 with NRSTS died ofthe disease after local relapse. We underline that the 4patients younger than 3 years, who underwent PRE, arealive and well (1 in second CR).

Primary reexcision, if feasible, represents the treat-ment of choice, in particular for children <3 years of agefor whom RT would increase important late sequelae.Our data suggest that PRE and RT are equally effectivein the treatment of NRSTS. Local treatment of micro-scopic residuals is highly recommended for these sarco-mas, which are generally less chemosensitive than RMS[14]. The feasibility of PRE depends on the site but alsoon the initial size of the tumor mass. In our group, PREwas not successful in 50% of patients whose mass ini-tially was >5 cm in size.

The efficacy of RT in improving local control in pa-tients with microscopic disease is widely accepted[1,2,3,15]. However, the timing and the modalities ofadministration follow different strategies. Whereas SIOPprotocols prefer to continue CT, reserving RT for pa-tients who relapse, the guidelines of the IRS and of otherEuropean studies recommend an early RT [16,17].

TABLE IV. Outcome According to Treatments and Histologies

Treatment

ALL (n 4 52) RMS (n4 25) NRSTS (n4 27)

Pts First CR LR (D)/MET (D) First CR LR/MET First CR LR/MET

PRE + CT 18 13 3 (1)/2 (2) 5 1/1 8 2/1RT + CTa 27 22 4 (1)/1 (1) 12 2/— 10 2/1CT 7 4 3 (3)/— 4 —/— — 3/—Total 52 39 10 (5)/3 (3) 21 3/1 18 7/2

CR, complete remission; D, died of disease; LR, local relapse; MET, metastases.aIn this group are included 3 cases in whom an initial PRE failed.

100 Cecchetto et al.

CONCLUSIONS

Our results show that microscopic residuals after ini-tial surgery are difficult to cure; they should be avoidedby an accurate diagnostic workup and by precise evalu-ation of the feasibility of primary excision. Local treat-ment with PRE or RT cured most of our patients withmicroscopic incomplete resection but did not always pre-vent local relapses. It was particularly important forNRSTS, which generally are less chemosensitive thanRMS.

PRE was the treatment of choice for paratesticulartumors and in children under 3 years of age, for whomRT may cause important sequelae. Also, extremity andtrunk sarcomas were cured by PRE; however, this wasnot effective for tumors of initial diameter >5 cm. RTachieved results similar to those with PRE in these lo-calizations. Thus the criteria that make PRE preferableinclude the age of the patient, the site, and the initial sizeof the tumor. Adequate RT was effective when micro-scopic residuals could not be avoided and PRE was notpossible. Although the disease-free survival after PRE isslightly lower than that observed after RT, local failuresare comparable in the two groups.

ACKNOWLEDGMENTS

The authors thank Ilaria Zanetti for statistical consul-tation and for her help in the research and Mara Fortin forsecretarial assistance.

REFERENCES

1. Carli M, Guglielmi M, Sotti G, et al. Soft tissue sarcomas. In:Plowman PN, Pinkerton CR, editors. Paediatric oncology, clinicalpractice and controversies, 2nd ed. London: Chapman and Hall;1996. p 380–416.

2. Andrassy RJ. Rhabdomyosarcoma. Semin Pediatr Surg 1997;6:17–23.

3. Flamant F, Rodary C, Rey A, et al. Treatment of non metastaticrhabdomyosarcomas in childhood and adolescence. Results of thesecond Study of the International Society of Paediatric Oncology.Eur J Cancer 1998;34:1050–1068.

4. Ennekin WF, Spanier SS, Malawer MM. The effect of anatomic

setting on the results of surgical procedures for soft parts sarcomaof the thigh. Cancer 1981;47:1005–1022.

5. Goodlad JR, Fletcher CDM, Smith MA, et al. Surgical resection ofprimary soft tissue sarcoma. Incidence of residual tumor in 95patients needing reexcision after local resection. J Bone Joint Surg[Br] 1996;78:658–661.

6. Zornig C, Peiper M, Schroder S. Reexcision of soft tissue sarco-mas after inadequate initial operation. Br J Surg 1995;82:278–279.

7. Davis AM, Kandel RA, Wumder JS, et al. The impact of residualdisease on local recurrence in patients treated by initial unplannedresection for soft tissue sarcoma of extremity. J Surg Oncol 1997;66:81–87.

8. Hays DM, Lawrence W, Wharam M, et al. Primary reexcision forpatients with microscopic residual tumor following initial excisionof sarcomas of trunk and extremity sites. J Pediatr Surg 1989;24:5–10.

9. Bell RS, O’Sullivan B, Liu F, et al. The surgical margins in softtissue sarcoma. J Bone Joint Surg [Am] 1989;71:370–375.

10. Tefft M, Lindberg RD, Gehan A. Radiation therapy combinedwith systemic chemotherapy of RMS in children: local control inpatients enrolled in the IRS. NCI Monogr 1981;56:75–81.

11. Donaldson S, Asmar L, Frier C, et al. Hyperfractionated radiationprotocol improve local control without increasing late effects inchildhood RMS. Med Pediatr Oncol 1994;23:188.

12. Oberlin O, Rodary C, Praquin MT, et al. Is it possible to cure RMSwithout local therapy? Results of the second study of the Interna-tional Society of Paediatric Oncology (MMT84) [abstract]. ThirdInternational Congress on soft tissue sarcoma in children and ado-lescent, Stuttgart, 1997. p 63.

13. Pilgrim TB, Klein S, Porstman S, et al. The value of chemotherapyin providing local tumor control in STS in children and adoles-cents. A retrospective analysis of CWS patients without localtherapy [abstract]. Third International Congress on soft tissue sar-coma in children and adolescent, Stuttgart, 1997. p 64.

14. Dillon PW. Non-rhabdomyosarcoma soft tissue sarcomas in chil-dren. Semin Pediatr Surg 1997;6:24–28.

15. Regine WF, Fontanesi J, Kumar P, et al: Local tumor control inrhabdomyosarcoma following low-dose irradiation: comparisonof Gr. II and select Gr. III patients. Int J Rad Oncol Biol Phys1995;31:485–491.

16. Habrand JL, Spooner D, Oberlin O, et al. The place of radiationtherapy in the management of localized MMT. The SIOP Expe-rience [abstract]. Third International Congress on soft tissue sar-coma in children and adolescent, Stuttgart, 1997. p 47.

17. Schmidt B.F, Herbst M, Potter R, et al. The role of RT in thetreatment of childood STS. Experience of the German CWSGroup [abstract]. Third International Congress on soft tissue sar-coma in children and adolescent, Stuttgart, 1997. p 49.

Microresidual Pediatric Soft Tissue Sarcoma 101