With an annual incidence of 4.3 cases per million persons age 20

Transcription

1 2597 Rhabdomyosarcoma in Infants Younger than One Year Old A Report from the Italian Cooperative Group Andrea Ferrari, M.D. 1 Michela Casanova, M.D. 1 Gianni Bisogno, M.D. 2 Ilaria Zanetti, 2 Giovanni Cecchetto, M.D. 3 Bruno De Bernardi, M.D. 4 Riccardo Riccardi, M.D. 5 Paolo Tamaro, M.D. 6 Cristina Meazza, M.D. 1 Rita Alaggio, M.D. 7 Vito Ninfo, M.D. 7 Modesto Carli, M.D. 2 1 Pediatric Oncology Unit, Istituto Nazionale per lo Studio e la Cura dei Tumori, Milan, Italy. 2 Department of Pediatric Hematology-Oncology, Università di Padova, Padua, Italy 3 Department of Pediatric Surgery, Università di Padova, Padua, Italy. 4 Department of Pediatric Oncology, Istituto Giannina Gaslini, Genoa, Italy. 5 Department of Pediatric Oncology, Università Cattolica Policlinico Gemelli, Rome, Italy. 6 Department of Pediatric Oncology, Università di Trieste, Trieste, Italy. 7 Pathology Unit, Università di Padova, Padua, Italy. Supported in part by Fondazione Città della Speranza and by Ministero dell Istruzione, dell Università e della Ricerca (MURST) ex 60%. Address for reprints: Andrea Ferrari, M.D., Pediatric Oncology Unit, Istituto Nazionale Tumori, Via G. Venezian, Milano MI, Italy; Fax: (011) ; andrea.ferrari@istitutotumori.mi.it Received December 31, 2002; revision received January 28, 2003; accepted January 28, BACKGROUND. The management of rhabdomyosarcoma (RMS) in patients age 1 year is particularly problematic and requires a tailored therapeutic approach. We report on the Italian Cooperative Group s 20-year study of 50 children with RMS who were age 1 year at diagnosis. METHODS. Patients were treated using multimodality therapeutic approaches that were based on three consecutive protocols. Chemotherapy was administered to all patients, with dosages calculated according to body weight; calculated doses subsequently were reduced by 33 50%. Radiotherapy was administered to 10 patients. RESULTS. With a median follow-up of 76 months, 5-year event-free survival and overall survival rates were 42.3% and 61.7%, respectively. Local recurrence was the major reason for treatment failure. In particular, the local recurrence rate was high in patients who warranted radiotherapy but received none due to their age. Completeness of surgery and nodal involvement were the most significant prognostic factors. After a suitable reduction in dosage was made, acute toxicity was no different from what has been observed in older children. The most relevant toxic event was cardiotoxic death in a newborn (n 1). CONCLUSIONS. The current study confirmed that the outcome for infants with RMS is less satisfactory than for older children and that infants with RMS require more careful monitoring and specific treatment guidelines. The absence of local control is the major cause of treatment failure; aggressive conservative surgery should be encouraged, but more radiotherapy may be advisable in selected cases. Intensive chemotherapy is essential; a 33% dose reduction may ensure adequate tolerance. In addition, patients age 3 months should not receive anthracyclines. Cancer 2003;97: American Cancer Society. DOI /cncr KEYWORDS: infants, soft tissue sarcomas, rhabdomyosarcoma, chemotherapy, cardiotoxicity. With an annual incidence of 4.3 cases per million persons age 20 years, rhabdomyosarcoma (RMS) is the most common soft tissue sarcoma in children. RMS can occur at any age; its peak incidence occurs at ages 2 6 years, with a second peak at ages years. 1 RMS may be present at birth: approximately 5 10% of all cases occur in patients age 1 year, and 1 2% of all cases are congenital. 2 Cooperative multidisciplinary treatment protocols have substantially improved the outcome of children with RMS over the past 25 years. Currently, patients with localized disease can be cured in 70% of all cases; 1,3 7 nonetheless, the small percentage of patients 2003 American Cancer Society

2 2598 CANCER May 15, 2003 / Volume 97 / Number 10 age 1 year are particularly difficult to manage. The well-known physiologic immaturity of various organs is responsible for the vulnerability of infants to acute and late effects of therapy, and the functional immaturity of the liver leads to a different metabolism of drugs in infants as compared with older patients. Infants with RMS require tailored treatments; the use of radiotherapy is restricted by the very high risk of side effects, and the morbidity of full-dose chemotherapy is unacceptable. Important points concerning treatment modalities in infants have yet to be clarified completely. Various reports indicate that the outcome for infants is the same as for older children and is not affected by dose reduction, 8 11 but more recent data suggest a less favorable outcome for patients age 1 year. 4,14 With the goal of pinpointing the best treatment modality for infant patients with RMS, we analyzed the 20-year study of the Italian Cooperative Group (ICG) on treatment of RMS in patients age 1 year. MATERIALS AND METHODS Fifty children with RMS who were age 1 year at diagnosis were treated between 1979 and 2001 in centers affiliated with the Associazione Italiana Ematologia Oncologia Pediatrica: 43 patients were enrolled on 1 of the 3 ICG protocols (10 on the RMS-79 protocol, 8 16 on RMS-88, and 17 on RMS-96), and 7 patients were treated through 1996 according to the protocols of the Istituto Nazionale Tumori (INT) of Milan (after 1996, patients in Milan were included in the ICG studies). This series of patients represented 6.6% (50 of 761) of all pediatric patients with RMS registered during the study period (10 of the 161 patients enrolled on the RMS-79 protocol, 16 of the 252 on RMS-88, 17 of the 175 on RMS-96, and 7 of the 173 on the INT protocol). Histopathologic data, clinical details, treatment modalities, and outcome were reviewed. Histologic categorization was based on the pediatric International Classification of Rhabdomyosarcoma. 13 Investigation at diagnosis included the following: physical examination, evaluation of local extent with computerized tomography (CT) and/or nuclear magnetic resonance imaging, staging with conventional chest X- ray or chest CT scan, abdominal ultrasound or abdominal CT scan, whole body bone scan, and bone marrow aspirates. Disease stage was assigned according to both the clinical TNM pretreatment staging system 14 and the Intergroup Rhabdomyosarcoma Study (IRS) postsurgical grouping system. 15 The TNM system defines T1 tumors as those that are confined to the organ or tissue of origin, whereas T2 lesions invade contiguous structures; the T1 and T2 groups are divided into A and B subgroups according to tumor diameter (A, 5 cm; B, 5 cm). Regional node involvement was classified as N0 (no lymph node involvement) or N1 (positive nodes) on the basis of clinical and radiologic assessment. For classification of distant metastases, M0 is defined as no metastases and M1 as metastases at onset. 14 After initial surgery, patients were classified according to the IRS system. Group I includes patients with completely excised tumors; Group II includes patients with grossly resected tumors, with microscopic residual disease and/or regional lymph node spread that has been resected completely; Group III includes patients with gross residual disease after incomplete resection or biopsy; and Group IV includes patients with metastases at diagnosis. 15 Patients were treated using multimodality therapeutic approaches including surgery, chemotherapy, and, in selected cases, radiotherapy. Treatment protocols varied during the study period, but overall strategies did not change substantially. Primary excision was attempted when complete and nonmutilating resection was considered feasible; otherwise, a biopsy was taken and chemotherapy was administered to shrink the tumor and make it resectable at subsequent surgery. Table 1 shows the different chemotherapy schedules used in the different protocols. The first Italian protocol and the INT protocol used regimens involving vincristine, doxorubicin (Adriamycin; Adria Laboratories, Columbus, OH), and cyclophosphamide (VAC/CAV) and vincristine, doxorubicin, cyclophosphamide, and actinomycin-d (VACA), respectively. Ifosfamide replaced cyclophosphamide in VAIA and IVA regimens in subsequent studies. Carboplatin and etoposide, along with vincristine, doxorubicin, and ifosfamide, were used in the CEVAIE schedule. Drug doses were calculated based on body weight, resulting in an approximately 30% reduction relative to doses that were calculated based on body surface area. Doses were calculated using 30 kg 1m 2 as the conversion factor. Doses then were reduced further (Table 1) to contain the risk of severe toxicities. In all ICG protocols, chemotherapy doses were increased gradually in subsequent courses if the first cycles were well tolerated. To minimize the burden of late sequelae, radiotherapy was used to treat only a small subgroup of children who had unresectable tumors and were unable to achieve complete tumor regression with chemotherapy. When feasible, irradiation was not administered between the 10th and 14th weeks as it is in older children; instead, it was deferred until after the

3 Rhabdomyosarcoma in Infants/Ferrari et al TABLE 1 Chemotherapy Regimens VAC/CAV (ICG RMS-79 protocol) VAC: Vincristine, 1.5 mg/m 2 i.v. (max, 2 mg) on Day 8; actinomycin-d, 1.5 mg/m 2 i.v. (max, 2 mg) on Day 8; cyclophosphamide, 150 mg/m 2 i.m. or orally on Days 1 7 CAV: Cyclophosphamide, 150 mg/m 2 i.m. or orally on Days 1 7; doxorubicin, 35 mg/m 2 i.v. on Day 8; vincristine, 1.5 mg/m 2 i.v. (max, 2 mg) on Day 8; for 52 wks VAIA (ICG RMS-88 and RMS-96 protocols) Vincristine, 1.5 mg/m 2 i.v. (max, 2 mg) weekly for 7 wks, then on Wk 9; actinomycin-d, 1.5 mg/m 2 i.v. (max, 2 mg) on Day 1 of Wks 1 and 7; ifosfamide 2 g/m 2 i.v. on Days 1 5 (or 3 g/m 2 i.v. 2 days) of Wks 1, 4, and 7; doxorubicin, 40 mg/m 2 i.v. on Days 1 and 2 of Wk 4; for 27 wks IVA (ICG RMS-88 and RMS-96 protocols) Ifosfamide, 3 g/m 2 i.v. 3 (or 2) days on Wks 1, 4, and 7; vincristine, 1.5 mg/m 2 i.v. (max, 2 mg) weekly for 7 wks, then on Wk 9; actinomycin-d, 1.5 mg/m 2 i.v. on Day 1 of Wk 7; for 27 wks VACA (INT protocol) Vincristine, 1.5 mg/m 2 i.v. (max, 2 mg) weekly for 7 wks, then on Wk 9; actinomycin-d, 0.5 mg/m 2 i.v. (max, 2 mg) on Days 1 3 of Wk 4; cyclophosphamide, 1200 mg/m 2 i.v. on Day 1 of Wks 1, 4, and 7; doxorubicin, 30 mg/m 2 i.v. on Days 1 and 2 of Wks 1 and 7; for 52, 37, or 26 wks CEVAIE (ICG RMS-96 protocol) Carboplatin, 500 mg/m 2 on Wk 1; epi-doxorubicin, 150 mg/m 2 on Wk 1; vincristine, 1.5 mg/m 2 weekly for 7 wks, then on Wk 9; actinomycin-d, 1.5 mg/m 2 on Wk 4; ifosfamide, 3 g/m 2 /d 3 days on Wks 4 and 7; etoposide, 150 mg/m 2 /d 3 days on Wk 7; for 27 wks Dose modifications ICG RMS-79 protocol: doses calculated according to body weight, then reduced by 50% for infants age 6mos ICG RMS-88 protocol: doses calculated according to body weight, then reduced by 50% for infants age 12 mos ICG RMS-96 protocol: doses calculated according to body weight, then reduced by 50% for infants age 6 mos; no anthracyclines for infants age 3 mos; anthracycline dose reduced by 33% for infants age 12 mos if weight 10 kg (full anthracycline doses if weight 10 kg) a All ICG protocols: doses increased in subsequent cycles if well tolerated INT protocol: doses calculated according to body weight, then reduced by 33% for infants age 12 mos ICG: Italian Cooperative Group; RMS: rhabdomyosarcoma; i.v.: intravenously; i.m.: intramuscularly; INT: Istituto Nazionale Tumori. a Anthracycline modifications for IVA regimen only. chemotherapy courses to facilitate maximum response to chemotherapy, to allow for possible delayed surgery on residual disease, and to let the child grow in the meantime. Radiotherapy was delivered by external beam radiation, with conventional fractionation ( cgy daily) for a total dose of Gy or with hyperfractionated accelerated radiotherapy (2 daily fractions of 160 cgy) for a dose of Gy. High-dose remote brachytherapy was considered as another option. Hyperfractionated radiotherapy and brachytherapy were used with the goal of reducing late effects. Administration of radiotherapy required immobilization and sedation. Response to treatment was evaluated at Weeks 9 and 18 and at the end of the entire treatment. Evaluation was based on the reduction in volume of all measurable lesions, and responses were defined as follows: complete response (CR) complete disappearance of disease; partial response (PR) tumor volume reduction by greater than two-thirds; minor response (MR) tumor volume reduction by greater than one-third. Stable disease or tumor volume reduction by less than one-third was recorded as no response, and an increase in tumor size or the detection of new lesions was recorded as progression of disease. Event-free survival (EFS) and overall survival (OS) were estimated according to the Kaplan Meier method. 18 For EFS, patients were evaluated from the date of diagnosis to the date of disease progression, recurrence, or death due to any cause. For OS, patients were evaluated from diagnosis to death. The time scale extended as far as the most recent followup if none of these endpoints were reached. To establish the potential value of prognostic factors, survival curves for different subgroups of patients were compared using the log rank test. Patient follow-up as of September 2002 ranged from months (median, 76 months). RESULTS Clinical characteristics of the 50 patients are shown in Table 2 and are compared with those of older children and adolescents enrolled on the same protocols during the study period. Of the 50 patients in the current study, 15 were considered as having congenital RMS: this subset included all patients diagnosed with RMS within 4 weeks of birth and those who later were diagnosed with a lesion that clearly was present at birth. No major differences were seen between infant patients and older patients. The most common primary site in infants was the nonorbital, nonparameningeal head and neck region. Only three infants had distant metastases at diagnosis.

4 2600 CANCER May 15, 2003 / Volume 97 / Number 10 TABLE 2 Comparison Between Current Series and Pediatric Patients Enrolled on the ICG RMS-79/88/96 and INT Protocols Infants a No. of patients Period Gender (M/F) 29/21 446/315 Age range (median) 1yr(6mos) 1 18 yrs (8 yrs) Histology (%) Embryonal b Alveolar Other c 4 2 Site of origin (%) Head-neck Parameningeal Orbital 4 8 HN non-pm Genitourinary Bladder-prostate GU non-bp 4 13 Extremities 10 8 Other d TNM stage (%) T N M Size e (%) 5cm IRS stage (%) I II III IV 6 12 ICG RMS-79/88/96 INT protocols ICG: Italian Cooperative Group; RMS: rhabdomyosarcoma; INT: Istituto Nazionale Tumori; HN: headneck; PM: parameningeal; GU: genitourinary; BP: bladder-prostate; IRS: Intergroup Rhabdomyasarcoma Studies; NOS: not otherwise specified. a Includes 15 patients with congenital RMS. b Includes two patients with botryoid RMS and two with spindle cell RMS. c Includes one patient with pleomorphic RMS and one with RMS NOS. d Includes five patients with trunk as site of origin, five with perineum as site of origin, and two with abdomen as site of origin. e Range, 1 12 cm; median, 4 cm. Treatment Sixteen patients (32%) had complete tumor resection. Five patients had complete resection at diagnosis (IRS Group I patients), and 11 patients had complete resection at surgery that was delayed until after primary chemotherapy. All surgical resections were conservative. Radiotherapy was administered to 10 patients (dose range, Gy; median, 44 Gy). External beam radiation was used in 8 cases (4 with conventional fractionation and 4 with hyperfractionated accelerated radiotherapy), and brachytherapy was used in 2 cases. Radiotherapy was administered to 4 patients with parameningeal RMS; to 1 child who had a large orbital tumor and responded poorly to chemotherapy, after a delayed incomplete resection; and to 5 patients with large tumors (2 in the head and neck region, 2 in the pelvis, and 1 on the left arm) that were considered unresectable at diagnosis. All patients received chemotherapy. In 80% of cases, chemotherapy regimens included anthracyclines. In compliance with the ongoing protocol guidelines, the majority of patients had a reduced dose; only 4 (8%) received full-dose chemotherapy. The distribution of treatment regimens was as follows: 10 VAC/CAV, 7 VACA, 14 VAIA, 10 IVA, and 9 CEVAIE. Response to chemotherapy at Week 9 was evaluable for 37 patients: 11 had CR, 14 had PR, and 7 had MR; the overall response rate was 86%. Five patients showed no response to first-line chemotherapy. Overall, 35 of 50 infants achieved CR during the treatment. Five patients achieved CR after initial surgery, 11 after chemotherapy and delayed surgery (5 of the 11 also received radiotherapy after gross resection), and 10 after incomplete surgery and chemotherapy. Three patients achieved CR after chemotherapy, surgery, and postoperative radiotherapy on residual tumor. In 6 patients, complete disappearance of the tumor was obtained with chemotherapy alone. The maximum tumor regression was a PR in 10 patients. Outcome With a median follow-up of 76 months, 5-year EFS and OS were 42.3% and 61.7%, respectively, and decreased to 39.2% and 57.9% at 10 years, due to a late metastatic failure (at 65 months). Twenty-two patients were alive in first CR, and six were alive in second CR; four patients were alive with disease. Seventeen patients died of progressive disease, and one patient died of acute cardiotoxicity (see below). Twenty-six patients experienced recurrence at 3 65 months (median, 9 months) after diagnosis. Treatment failure was local in 85% of patients. The sites of recurrence were as follows: 18 local, 2 local and nodal, 2 local and distant, 2 regional lymph node, and 2 distant metastases. Of the patients who experienced recurrence, 17 died of disease at 4 87 months after diagnosis, despite additional therapy. Six patients (4 with local recurrence and 2 with regional recurrence) were in second CR after secondary treatment that included chemotherapy plus surgery and/or radiotherapy. Of the 18 patients who experienced local failure, 4 (22%) were alive in second CR at months after recurrence, 12 died of disease, and 2 were alive with disease at the time of the current analysis.

5 Rhabdomyosarcoma in Infants/Ferrari et al TABLE 3 Univariate Analysis According to Clinical Findings 5-year EFS a (%) P Male (n 29) vs. female (n 21) 42 vs Congenital (n 15) vs. noncongenital (n 35) 44 vs Alveolar (n 11) vs. nonalveolar (n 38) 27 vs T1 (n 26) vs. T2 (n 24) 48 vs N0 (n 42) vs. N1 (n 8) 49 vs Tumor size 5cm(n 28) vs. 5cm(n 22) 47 vs IRS Group I (n 5) vs. II (n 8) vs. III (n 34) vs. IV (n 3) 100 vs. 50 vs. 36 vs ICG RMS-79 protocol (n 10) vs. RMS-88 (n 16) vs. RMS-96 (n 17) 54 vs. 33 vs EFS: event-free survival; IRS: Intergroup Rhabdomyosarcoma Studies; ICG: Italian Cooperative Group; RMS: rhabdomyosarcoma. a Five-year EFS 42.3% for all patients. The local failure rate was particularly high for patients whose stage of disease warranted radiotherapy according to the protocol guidelines for older children but who received none because of their age (14 of 26 [54%]), compared with patients who did not receive radiotherapy according to protocol (4 of 14 [28%]). The local failure rate also was high for patients who were treated with radiotherapy (4 of 10 [40%]), but children in this subset had a high risk of local failure, as they were treated for invasive and large unresectable tumors. Table 3 shows the univariate analysis comparing the estimated EFS for different subsets of patients stratified according to various clinical features. Given the small sample size, the only statistically significant difference in outcome was related to nodal involvement. With regard to histology, the EFS clearly was better for patients with nonalveolar RMS than for patients with alveolar RMS, but probably due to the small number of alveolar cases the difference was not significant. EFS also was related to IRS grouping; the role of surgical resection became evident when we considered the number of recurrences among patients who underwent complete resection at first approach or at delayed surgery (4 recurrences among 15 patients [26%]) compared with those who did not (22 recurrences among 35 patients [63%]). We observed no differences in outcome related to tumor site. Our analysis showed a distinctly worse EFS for patients enrolled on the RMS-88 protocol compared with those on RMS-79 or RMS-96. This finding appeared to be due to the clinical features of the patients enrolled on the different protocols. Clinical presentations were less favorable among patients enrolled on the RMS-88 protocol. In the RMS-79 group (n 10), 3 children were classified as IRS Group I (as opposed to 1 in each of the other 2 protocols), none had nodal or distant metastases at diagnosis, and 2 had alveolar histologies; in the RMS-88 group (n 10), 5 patients were classified as N1 and 2 as M1, and 5 patients had alveolar RMS; in the RMS-96 group (n 17), 2 patients were classified as N1 and 1 as M1, and 3 patients had alveolar RMS. Toxicity The most relevant toxic event that occurred in this series was a death caused by acute cardiotoxicity. The patient who died was treated in 1995 and was very young at diagnosis (age, 3 weeks). The patient died of heart failure within a few days of the administration of a second cycle of VAIA chemotherapy. In the chemotherapy regimen for this patient, doxorubicin, which is scheduled for older patients at 40 mg/m 2 (6-hour infusion) for 2 days, was administered at a reduced dose of 0.7 mg/kg (the dose calculated according to body weight and reduced by 50%) for 2 days. Outside the present series, a similar event occurred in a newborn who, according to the Italian RMS-88 protocol, was treated for a non-rms soft tissue sarcoma using the same chemotherapy regimen. No significant hematologic, renal, hepatic, or neurologic toxicities were reported as a consequence of chemotherapy. Neutropenia generally was mild, and no severe infections occurred. To avoid severe reactions (e.g., mucositis in patients with head and neck RMS) related to the concomitant administration of chemotherapy and radiotherapy, patients receiving radiotherapy had their treatment delayed until chemotherapy was completed. Local reactions did not differ from those observed in older children. Six patients who received radiotherapy were longterm survivors; as expected, significant bone and soft tissue growth retardation was evident in all six. No data regarding any potential endocrine dysfunction were available. None of the six patients developed second malignancies.

6 2602 CANCER May 15, 2003 / Volume 97 / Number 10 DISCUSSION The clinical management of malignancies in patients age 1 year is particularly difficult and requires special therapy modifications and careful monitoring to prevent severe (sometimes life-threatening) complications and disabling late sequelae. 17 The immaturity of various organs results in the particular susceptibility of infants to side effects and leads to qualitative and quantitative differences in drug metabolism and pharmacology that still are not understood fully, especially in newborns. Renal clearance of drugs is slower for infants than for older children; consequently, prolonged plasma half-life and sometimes unpredictable changes in hepatic drug metabolism and agent binding by plasma proteins are observed. In addition, there are known differences in body water volume (80% of body weight in neonates vs. 50% in older children). 17 Several studies of infants with RMS have attempted to identify the optimal treatment strategy. The report of Ragab et al. 9 on infants treated in the IRS-I and IRS-II trials revealed unacceptable toxicity when full chemotherapy doses were used (5% of treatment-related deaths in infants vs. 1% in older children). This finding prompted an amendment to the IRS-II protocol that called for a 50% dose reduction; the amendment resulted in a clear decrease in lifethreatening and fatal toxicities, while the overall outcome did not differ from the outcome for older children (survival was worse only for Group III patients, due to an increase in the local failure rate). 9 Reports from the ICG RMS-79 study (n 10), 8 the Institut Gustave-Roussy (n 43) study, 11 and the German CWS-81 study (n 14) 10 confirmed that dose reduction in infants was tolerable and did not affect the survival rate. A recent report from the International Society of Pediatric Oncology (SIOP) on 102 infants with malignant mesenchymal tumors (64 of which were RMS) described a satisfactory overall outcome: chemotherapy was manageable with appropriate dose reductions, and poor results were observed only in patients with alveolar RMS or metastases at onset. 18 More recently, age 1 year emerged as an independent adverse prognostic factor for RMS (as did age 10 years). Data from the Italian RMS-88 study were consistent with the IRS analyses: in the IRS-IV study, 4 failure-free survival was 55% for infants, 83% for children age 1 9 years, and 68% for patients age 10 years. Joshi et al. 12 described the difficulties in clearly identifying the reasons for the difference in outcomes. The results might have been influenced by the subset of patients diagnosed in the neonatal period; these patients were recognized as having a worse outcome in numerous studies The IRS group reported a 3-year OS of 49% among 14 newborns with RMS, 20 and the Children s Cancer Group reported 9 treatment failures among 11 patients with congenital RMS. 19 In the current series, no major differences in clinical characteristics or histologic pattern were observed between infants and older children. We did not confirm the findings of others regarding a high frequency of botryoid and undifferentiated histologies. Most patients had tumors that were considered unresectable at diagnosis, while the proportion of metastatic cases (6%) was lower than expected. The treatment plans adopted for infants were the same as those used for older children, except for the drug dose modifications and the indications for radiotherapy. (Radiotherapy was omitted in most cases and otherwise was delayed until the end of treatment.) The current study included patients who were treated during a 20-year period. The OS rate for infants (61.7%) was lower than for older children (5-year OS, 67.2% for pediatric patients enrolled on the ICG RMS-88 protocol). The difference between infants and older children with RMS was demonstrated more clearly by EFS; because radiotherapy was used less often to treat younger children, their rate of local recurrence was higher, although a subset of patients with local recurrence were cured with salvage therapy and adequate local treatment (surgery and/or radiotherapy). Lack of local control was the primary cause of treatment failure. The incidence of local recurrence was especially high in patients whose situations (i.e., incomplete resection or alveolar histologies) warranted radiotherapy but who did not receive radiotherapy due to their age. Although every effort should be made to encourage and optimize treatment plans that include chemotherapy plus aggressive but conservative complete surgery and that limit the use of radiotherapy, our findings nonetheless suggest that increased use of radiotherapy may be appropriate in selected cases. Alternative techniques (e.g., three-dimensional conformal radiotherapy, hyperfractionated accelerated radiotherapy, and brachytherapy) capable of minimizing the dose received by normal surrounding tissues (and thereby limiting sequelae) should be considered. As reported elsewhere, 8 14 reduced-dose chemotherapy was well tolerated in the current series. No major renal or hepatic toxicity or severe neurotoxicity secondary to vincristine (which reportedly is more

7 Rhabdomyosarcoma in Infants/Ferrari et al toxic in infants, particularly newborns) was observed, and only moderate myelosuppression was observed. The one case of death due to heart failure was similar to a case observed in another patient with non-rms soft tissue sarcoma; both were newborn, and both received reduced-dose doxorubicin. The risk of cardiotoxicity following anthracycline treatment is well known; in a large review by the Pediatric Oncology Group, cardiotoxicity was observed in 1.6% of all patients. 23 Although it is difficult to estimate the risk of death due to heart failure, and although young age usually is not considered a risk factor for anthracycline-associated cardiotoxicity, 24,25 our data may suggest that anthracyclines, even at lower doses, should be avoided in children age 3 months. Calculating drug dosage on the basis of surface area probably leads to an overestimation in infants, so doses usually are calculated according to body weight instead. However, the absence of any other severe acute toxicities (particularly neutropenia) in the current series may indicate that a 50% reduction of the dose calculated by body weight may be excessive. One option is to recommend a 33% dose reduction for children age 6 months; thereafter, as long as it is well tolerated, the dose could be increased until full dose is reached. In conclusion, the current report sheds more light on the management of RMS in patients age 1 year. Careful monitoring and tailored treatment guidelines are needed in these patients to limit therapy-related morbidity while achieving the same results as in older children. With appropriate dose reductions, chemotherapy is manageable and effective. A double reduction (i.e., calculating dosage according to body weight rather than area, followed by an additional reduction) probably is unnecessary. However, particular care should be taken with regard to administration of anthracyclines in newborns; we recommend avoiding them in patients age 3 months. The first challenge in treating RMS in infants is improving local control. Aggressive conservative surgery is essential, but so is greater use of radiotherapy in selected cases (e.g., alveolar cases). The substantial improvement in outcome for children with RMS over the past 25 years seems to be related closely to the increased intensity of combined treatment programs, especially for high-risk patients. If infants are determined to be high-risk patients, more intensive treatment could be warranted. REFERENCES 1. Wexler LH, Crist WM, Helman LJ. Rhabdomyosarcoma and the undifferentiated sarcomas. In: Pizzo PA, Poplack DG, editors. Principles and practice of pediatric oncology, 4th edition. Philadelphia: Lippincott Williams & Wilkins, 2002: Weiss SW, Goldblum JR. Rhabdomyosarcoma. In: Weiss SW, Goldblum JR, editors. Enzinger and Weiss s soft tissue tumors (4th edition). St Louis: CV Mosby, 2001: Crist WM, Garnsey L, Beltangady MS, et al. Prognosis in children with rhabdomyosarcoma: a report of the Intergroup Rhabdomyosarcoma Studies I and II. J Clin Oncol. 1990;8: Crist WM, Anderson JR, Meza JL, et al. Intergroup Rhabdomyosarcoma Study-IV: results for patients with nonmetastatic disease. J Clin Oncol. 2001;19: Raney RB, Anderson JR, Barr FG, et al. Rhabdomyosarcoma and undifferentiated sarcoma in the first two decades of life: a selective review of Intergroup Rhabdomyosarcoma Study Group experience and rationale for Intergroup Rhabdomyosarcoma Study V. J Pediatr Hematol Oncol. 2001;23: Flamant F, Rodary C, Rey A, et al. Treatment of non-metastatic rhabdomyosarcomas in childhood and adolescence. Results of the second study of the International Society of Paediatric Oncology: MMT84. Eur J Cancer. 1998;34: Koscielniak E, Harms D, Henze G, et al. Results of treatment for soft tissue sarcoma in childhood and adolescence: a final report of the German Cooperative Soft Tissue Sarcoma Study CWS-86. J Clin Oncol. 1999;17: Carli M, Grotto P, Cordero di Montezemolo L, et al. Soft tissue sarcomas in infants less than 1 year old: experience of the Italian Cooperative Study RMS-79. Contrib Oncol Basel Karger. 1990,41: Ragab AH, Heyn R, Tefft M, Hays DN, Newton WA, Beltangady M. Infants younger than one year of age with rhabdomyosarcoma. Cancer. 1986;58: Koscielniak E, Harms D, Schmidt D, et al. Soft tissue sarcomas in infants younger than 1 year of age: a report of the German Soft Tossue Sarcoma Study Group (CWS-81). Med Pediatr Oncol. 1989;17: Salloum E, Flamant F, Rey A, et al. Rhabdomyosarcoma in infants under one year of age: experience of the Institut Gustave-Roussy. Med Pediatr Oncol. 1989;17: Joshi D, Anderson J, Paidas C, et al. Age is an independent prognostic factor in rhabdomyosarcoma (RMS): a report from the Intergroup Rhabdomyosarcoma Study group (IRS) (abstract 113). International Society of Paediatric Oncology XXXI Meeting, Montreal, Canada, September Med Pediatr Oncol. 1999;33: Newton WA, Gehan EA, Webber BL, et al. Classification of rhabdomyosarcomas and related sarcomas. Pathologic aspects and proposal for a new classification an Intergroup Rhabdomyosarcoma Study. Cancer. 1995;70: Harmer MH. TNM classification of pediatric tumors. Geneva: UICC International Union Against Cancer, 1982: Maurer HM, Beltangady M, Gehan EA, et al. The Intergroup Rhabdomyosarcoma Study I: a final report. Cancer. 1988;61: Kaplan E, Meier P. Non-parametric estimation from incomplete observation. J Am Stat Assoc. 1958;53: Morgan E, Baum E, Breslow N, Tahashima J, D Angio G. Chemotherapy-related toxicity in infants treated according to the Second National Wilms Tumor Study. J Clin Oncol. 1988;6:51 55.

8 2604 CANCER May 15, 2003 / Volume 97 / Number Orbach D, Rey A, Quintana E, Oberlin O, Sanchez de Toledo J, Stevens MCG. Malignant mesenchymal tumours (MMT) in infants [abstract O71]. International Society of Paediatric Oncology XXXIII Meeting, Brisbane, Australia, October Med Pediatr Oncol. 2001;37: Dillon PW, Whalen TV, Azizkhan RG, et al. Neonatal soft tissue sarcomas: the influence of pathology on treatment and survival. Children s Cancer Group Surgical Committee. J Pediatr Surg. 1995;30: Lobe TE, Wiener ES, Hays DM, et al. Neonatal rhabdomyosarcoma: the IRS experience. J Pediatr Surg. 1994;29: Siegel SE, Moran RG. Problems in the chemotherapy of cancer in the neonate. Am J Pediatr Hematol Oncol. 1981;3: Parkes SE, Muir KR, Southern L, Cameron AH, Darbyshire PJ, Stevens MCG. Neonatal tumours: a thirty-year population-based study. Med Pediatr Oncol. 1994;22: Krischer JP, Epstein S, Cuthbertson DD, Goorin AM, Epstein M, Lipshultz SE. Clinical cardiotoxicity following anthracycline treatment for childhood cancer: the Pediatric Oncology Group experience. J Clin Oncol. 1997;15: Grenier MA, Lipshultz SE. Epidemiology of anthracycline cardiotoxicity in children and adults. Semin Oncol. 1998;25: Marx M, Langer T, Graf N, et al. Multicentre analysis of anthracycline-induced cardiotoxicity in children following treatment according to the nephroblastoma studies SIOP No.9/GPOH and SIOP 93-01/GPOH. Med Pediatr Oncol. 2002;39:18 24.