Rhabdomyosarcoma in adults

A retrospective analysis of 171 patients treated at a single institution




The goal of the current study was to clarify treatment outcomes for adult patients with rhabdomyosarcoma (RMS). Published series have reported definitively worse results for adults with RMS compared with children with RMS. This finding casts doubt on whether RMS is the same disease in adults as it is in children.


Of 190 patients with RMS who were age 18 years or older and whose cases were recorded over a 25-year span in the pathology database of the Istituto Nazionale Tumori (Milan, Italy), 171 could be analyzed retrospectively for treatment outcome. The authors attempted to stratify patients according to the degree to which they had been treated appropriately, based on current treatment guidelines for childhood RMS.


The overall rate of response to chemotherapy was 85%. For the entire series, 5-year event-free survival and 5-year overall survival (OS) were 28% and 40%, respectively. Among the 110 patients with embryonal, alveolar, or ‘not otherwise specified’ RMS, 5-year OS was 46%; however, 5-year OS was 61% for patients within this group (39% of the total) who had high scores for appropriate treatment.


The current series parallels other published series in that it confirms the finding of a relatively poor long-term outcome for adult patients with RMS. However, for patients whose treatment adhered to the current guidelines for treatment of children, outcome was similar to what has been reported in pediatric series. In addition, the rate of response to chemotherapy for the entire series was similar to the rate typically observed among children. These findings suggest that adults and children with RMS should receive similar treatment. Treatment protocols adopted from pediatric programs but tailored to adults could increase adults' chances of receiving appropriate treatment; prospective studies are needed to test this idea. Cancer 2003;98:571–80. © 2003 American Cancer Society.

DOI 10.1002/cncr.11550

Rhabdomyosarcoma (RMS) is a rare malignancy. Nonetheless, it is a common childhood cancer, constituting more than 50% of all soft tissue sarcomas. In contrast, RMS is exceedingly infrequent in adults: soft tissue sarcomas make up less than 1% of all adult malignancies, and RMS accounts for 3% of all soft tissue sarcomas.1

There is no doubt that childhood RMS is a distinct entity. In particular, it differs from typical adult soft tissue sarcomas in terms of its natural history and its higher sensitivity to chemotherapy and radiotherapy. Currently, more than 70% of children with localized RMS can be cured with multidisciplinary treatment protocols that include chemotherapy.2 In contrast, standard treatment for adults with localized soft tissue sarcoma is based on surgery, often complemented by radiotherapy. To date, some benefit resulting from adjuvant chemotherapy has been demonstrated in adults with soft tissue sarcoma, but the level of benefit is not as high as it is for children with RMS.3 It therefore is reasonable to wonder what prognosis RMS has in adults and to what degree multidisciplinary approaches used to treat children are effective in adults. It is unfortunate that due to its rarity, scant data are available on clinical and biologic findings regarding adult RMS. Nonetheless, all existing studies report a poorer outcome for adults compared with children, despite the extrapolation of multimodality treatment from pediatric experience.

To augment existing data and assist in clarifying the issue of the applicability of childhood protocols to adults, we retrospectively analyzed the outcomes of 171 adult patients with diagnoses of RMS who were seen at the Istituto Nazionale Tumori (INT), Milan, Italy, over a 25-year period.


The cases of 190 consecutive patients age 18 years or older who were diagnosed with RMS between 1975 and 2001 were collected from the institutional database of the Pathology Unit of the INT. At pathology consultation, initial diagnoses of RMS were changed in 10 cases; 2 were changed to desmoplastic small round cell tumor, 2 to malignant peripheral nerve sheath tumor, and 1 each to liposarcoma, leiomyosarcoma, malignant hemangiopericytoma, clear cell sarcoma of soft parts, malignant fibrous histiocytoma, and melanoma. Consequently, 180 patients were available for the current analysis. However, data regarding treatment and outcome were incomplete in 9 cases, leaving 171 patients for the analysis. Patient age ranged from 19 to 83 years, with a median of 27 years. Most patients (n = 103) were younger than age 30 years, while 51 were between ages 31 and 60 years and 26 were older than age 61 years. All patients were white Caucasian.

In all cases, pathologic diagnosis was made before the start of treatment by pathologists at our institution, according to standard diagnostic criteria.4, 5 The botryoid variant of RMS was included in the embryonal subtype for the current analysis.

Clinical Grouping

In most cases, staging at diagnosis involved physical examination; evaluation of local extent with computerized tomography (CT) and/or magnetic resonance imaging; chest X-ray and/or chest CT scan; abdominal ultrasound or CT scan; whole-body bone scan; and, in patients with parameningeal primary site, cerebrospinal fluid cytology. In cases of RMS arising in the paratesticular region or the lower limbs, abdominal CT scans generally were available. Bilateral bone marrow aspirates and/or biopsies were performed in several cases of embryonal and alveolar RMS.

We attempted to retrospectively define stage of disease according to both the clinical TNM classification6 and the Intergroup Rhabdomyosarcoma Study (IRS) postsurgical grouping system.7 According to the TNM classification, T1 lesions are tumors confined to the organ or tissue of origin, whereas T2 lesions invade contiguous structures. The T1 and T2 groups are further divided into A and B subgroups, depending on whether tumor diameter is ≤ 5 cm (A) or > 5 cm (B). Regional node involvement was classified as N0 or N1, and distant metastases were classified at onset as M0 or M1, based on histologic or clinical/radiologic assessment.6 The IRS system categorizes patients into one of four groups based on the amount and extent of residual tumor after the initial surgical procedure: Group I includes completely excised tumors with negative microscopic margins; Group II includes macroscopically resected tumors with microscopic residual disease and/or regional lymph node spread; Group III includes patients with macroscopic residual disease after incomplete resection or biopsy; and Group IV includes patients with metastases at onset.7


Over a period of time spanning more than 25 years, patients were treated with a relatively consistent approach, which in most cases included surgery, chemotherapy, and radiotherapy. To compare treatment data and results with those from pediatric series, common definitions from the pediatric series were used. Surgical resection was defined as complete when histologic margins were free, and complete resection therefore included compartment resection (i.e., en bloc resection of the tumor and the entire compartment of origin)8 and wide excision (i.e., en bloc excision beyond the reactive zone but within the anatomic compartment).

Radiotherapy was delivered using photon or electron beam energies on the order of 1 megavolt, with conventional fractionation (1.8–2.0 grays [Gy] daily for 5 days per week). Overall, 109 patients received radiotherapy. The total dose ranged from 35 to 66 Gy, with a median dose of 54 Gy. With regard to timing, radiation was delivered within the first 12 weeks of treatment in 56% of all cases. In most cases, radiation fields included the initial volume of the tumor plus 2–3 cm margins, as well as any involved lymph nodes.

Chemotherapy was administered to 124 of 171 patients (72.5%). Different chemotherapeutic regimens were administered over the years covered by the study, according to ongoing protocols and usage at a given time. Most patients (n = 104) received a multidrug regimen that included cyclophosphamide or ifosfamide, in addition to doxorubicin, epirubicin, or dactinomycin; most regimens also included vincristine. Some patients also received dacarbazine, cisplatin, carboplatin, or etoposide in addition to cyclophosphamide/ifosfamide and anthracycline in various combinations or alternating regimens. In addition, five patients received intrathecal methotrexate for parameningeal RMS, and four patients with metastases received high-dose chemotherapy. In 20 cases, chemotherapy did not include cyclophosphamide or ifosfamide (i.e., single-agent chemotherapy with doxorubicin or a 2-agent regimen involving doxorubicin plus dacarbazine was used; both treatments were based on regimens used to treat adults with soft tissue sarcomas).

In the current analysis, response to chemotherapy was classified according to the radiologically assessed reduction in the sum of the products of the perpendicular diameters of all measurable lesions. Complete response (CR) was defined as the complete disappearance of disease, partial response (PR) was defined as a tumor reduction of > 50%, and minor response (MR) was defined as a maximum tumor reduction of > 25%. Stable disease or a reduction of < 25% was classified as no response, and an increase in tumor size or the detection of new lesions was classified as progression of disease.

Treatment Score

To retrospectively assess the prognosis of the patients in the current study while correcting outcomes for the obvious heterogeneity of treatments used over a period of more than 25 years, we arbitrarily defined a simple scoring system, based on current principles of treatment of childhood RMS, to categorize the adequacy of each patient's treatment. The current principles partially differ from those that were in use at the time patients were treated; the idea of dealing with adults with RMS separately from adults with soft tissue sarcomas has been appreciated fully by the medical community only in recent years.

Three scores were assigned to each patient—one for the adequacy of local treatment, one for the adequacy of chemotherapy, and one for the adequacy of the therapeutic strategy as a whole. Scores were consistent with the general strategy of the ongoing protocols adopted by the North American IRS group9, 10 and the Italian Cooperative Group (ICG). All 3 scores were multiplied together to yield a single number, which ranged from 0 to 1. In principle, an overall score of 1 corresponded to a treatment regimen that was entirely consistent with current recommendations for treatment of pediatric RMS.

With regard to local treatment, a score of 1 was assigned in cases of complete surgical resection, as defined above (with or without radiotherapy); and in cases of incomplete resection (with microscopically involved margins or macroscopic residual tumor) or biopsy, provided that these procedures were followed by adequate radiotherapy. Radiotherapy was considered adequate when the total tumor dose was > 50 Gy and was administered within 12–14 weeks of the surgical procedure. Radiotherapy at a total dose < 45 Gy (after incomplete resection) corresponded to a score of 0.6. Radiotherapy administered at an adequate dose more than 4 months after incomplete resection or biopsy corresponded to a score of 0.8. In cases of omission of radiotherapy after incomplete surgery, a score of 0 was given. The exception was a single case in which a patient experienced CR following chemotherapy and did not receive radiotherapy: a score of 0.7 was given, despite the lack of a consensus regarding this situation in the pediatric oncology community. In the statistical analysis, the local treatment score was cross-checked with local recurrence–free survival (LRFS) and overall survival (OS).

With regard to chemotherapy, a score of 0 was assigned to all cases in which chemotherapy was not used. A score of 1 was given if a multidrug regimen was incorporated into the treatment program; included cyclophosphamide or ifosfamide, as well as doxorubicin, epirubicin, and/or dactinomycin, with or without vincristine (and with or without other drugs, such as dacarbazine, cisplatin, carboplatin, and etoposide); and lasted for 8 cycles or more. We assigned scores of 0.4, 0.5, and 0.6 to chemotherapy regimens lasting 2, 4, and 6 cycles, respectively, and a score of 0.5 to chemotherapy regimens that did not include cyclophosphamide or ifosfamide (i.e., doxorubicin as a single agent and doxorubicin plus dacarbazine). We were unable to retrospectively assess the dose intensity of the regimens used. In the statistical analysis, the chemotherapy score was cross-checked with metastatic recurrence–free survival (MRFS) and OS.

For the chemotherapy analysis only, we excluded patients with pleomorphic RMS, because pleomorphic RMS is very rare in childhood and appears to be more similar, both biologically and clinically, to adult non-RMS soft tissue sarcomas. Consequently, no widely accepted guidelines are available for this subgroup, particularly regarding the role of chemotherapy.

Statistical Methods

EFS and OS were calculated according to the Kaplan–Meier method.11 Survival was evaluated from the date of histologic diagnosis before the first definite treatment to an event-free final follow-up, or to disease progression or recurrence (for EFS only) or death due to any cause (for both EFS and OS). LRFS was calculated from the time of diagnosis to local progression or recurrence (i.e., persistence or regrowth of tumor at the primary site). MRFS was calculated from the time of diagnosis to the development of distant metastases. The log rank test12 was used to compare the survival curves of the patient subgroups in a univariate analysis to ascertain the potential value of various prognostic factors. Follow-up, as of September 2002, ranged from 8 to 260 months (median, 50 months).


One hundred eighty patients were considered for the current study: 149 had localized disease, and 31 had metastases at diagnosis. Table 1 shows the clinical characteristics of the patients in the study. Histology was alveolar in 62 cases, embryonal in 60, pleomorphic in 37, and not otherwise specified (NOS) in 21. The head-neck region was the most common primary site. Regarding TNM classification, 64% of patients had T2b primary tumors, and 37% had lymph node involvement at diagnosis (N1). The primary tumor was larger than 5 cm in 77% of all cases.

Table 1. Patient Characteristics
  • IRS: Intergroup Rhabdomyosarcoma Study.

  • a

    Parameningeal in 34 cases and orbital in 4 cases.

  • b

    Lower extremities in 36 cases.

  • c

    Paratesticular in 20 cases.

  • d

    Intraabdominal in 24 cases, trunk in 16 cases, and mediastinal in 3 cases.

Study period1975–2001
No. of patients180
 Median (yrs)27
 Range (yrs)19–83
 19–30 yrs103
 31–60 yrs51
 > 60 yrs26
Disease spread 
 Not otherwise specified21
Site of origin 
TNM classification 
Size (cm) 
 ≤ 541
 > 5139
IRS group 

Data regarding treatment and outcome were complete for 171 patients. For these patients, 5-year EFS and OS were 27.9% and 39.6%, respectively. Patients with nonmetastatic embryonal, alveolar, or NOS histotype were analyzed separately from patients with pleomorphic RMS and patients with metastatic disease, as described below.

Embryonal, Alveolar, and NOS RMS—Localized Disease

The group of patients with embryonal, alveolar, or NOS RMS included 110 patients with complete data regarding clinical findings, treatment, and outcome. The median age in this group was 25 years, and only 8 patients were older than age 60 years. Histology was embryonal in 44 cases, alveolar in 48, and NOS in 18. Tumor size was less than 5 cm in 33 cases. According to the IRS classification system, 22 patients had Group I RMS, 22 had Group II RMS, and 66 had Group III RMS.

Complete surgical tumor resection was performed in 33 cases (30%): 22 patients with Group I RMS underwent complete resection as their first treatment (at onset in 15 cases and via reexcision within 8 weeks of an initial attempt at surgery in 7 cases), and 11 patients underwent complete tumor resection via delayed surgery, after tumor shrinkage induced by chemotherapy. Amputation was performed for four patients.

Radiotherapy was administered to 73 patients (66%). Of the 37 patients who did not receive radiotherapy as front-line treatment, 5 did not do so because of early progression and 2 refused treatment.

Regarding the local treatment score, 72 patients received a score of 1, 26 received a score of 0.8 (due to delay in the timing of irradiation), 1 received a score of 0.7, 6 received a score of 0.6, and 5 received a score of 0. Patients with a score of 1 had better LRFS and OS rates than did patients with lower scores (Table 2).

Table 2. Localized Embryonal, Alveolar, and ‘Not Otherwise Specified’ Rhabdomyosarcomas (n = 110): 5-Year Survival Rates by Treatment Score
 No. of patients5-year LRFS/MRFS/EFSa (%)5-year OS (%)b
  • LRFS: local recurrence–free survival; MRFS: metastatic recurrence–free survival; EFS: event-free survival.

  • a

    LRFS for local treatment score analysis, MRFS for chemotherapy score analysis, and EFS for combined score analysis.

  • b

    Five-year OS = 45.7% for entire cohort of 110 patients.

  • c

    Five-year LRFS = 59.5% for entire cohort of 110 patients.

  • d

    Five-year MRFS = 56.2% for entire cohort of 110 patients.

  • e

    Five-year EFS = 32.9% for entire cohort of 110 patients.

Local treatment scorec   
Chemotherapy scored   
Combined scoree   

Chemotherapy was administered to 88 patients (80%); 74 received a chemotherapy score of 1. Twenty-three patients did not receive systemic treatment (score, 0); 2 of these 23 patients refused treatment. Thirteen patients were assigned scores ranging from 0.4 to 0.6, due to their regimens or the duration of treatment. Patients who received adequate chemotherapy (score, 1) had a relatively good MRFS (Table 2). Of the 16 patients with recurrences who did not receive chemotherapy as their first treatment (score, 0), 5 were alive without disease after salvage chemotherapy (MRFS, 39.4%; OS, 49.4%).

Data regarding response to chemotherapy were available for 59 patients. Seven patients had CR, 40 had PR, and 3 had MR, for an overall response rate of 85%. Only nine patients did not respond to front-line chemotherapy.

Length of follow-up ranged from 8 to 260 months (median, 50 months). Five-year EFS and OS were 32.9% and 45.7%, respectively. At the time of the analysis, 38 patients were alive after achieving first CR, 10 after achieving a second CR, and 1 after achieving a third CR. Ten patients were alive with disease. One patient developed a second tumor (breast carcinoma) during CR, 10 years after diagnosis of RMS. Two patients died of treatment-related toxicity (sepsis [n = 1] and intracranial hemorrhage [n = 1]), and 48 died of progressive disease. Recurrence or progression of disease was local in 28 cases, local and nodal/distant in 8, nodal in 9, and distant in 22. Time to recurrence ranged from 2 to 144 months (median, 9 months).

Table 2 shows outcome according to treatment score for the current study population. Forty-three patients (39%) received fully adequate treatment and were assigned a score of 1. The percentage of patients with a score of 1 varied across age groups: among patients ages 19–30 years, 45% had a score of 1, compared with 29% of patients older than age 30 years. For the cohort of patients with a score of 1, outcome (5-year OS, 61.5%) was decidedly better than for other subgroups (Fig. 1). The small subset of 16 patients with both embryonal histotype and a score of 1 had a 5-year OS of 72.5%.

Figure 1.

Overall survival according to treatment score (includes patients with localized embryonal, alveolar, or ‘not otherwise specified’ rhabdomyosarcoma).

Table 3 shows the estimated EFS and OS rates of patients stratified according to various clinical characteristics. Patients with noninvasive tumors, patients with small tumors, and patients who received complete surgery at diagnosis (IRS Group I) had the best reported outcomes. Lymph node involvement and alveolar histology were associated with the worst reported outcomes.

Table 3. Localized Embryonal, Alveolar, and ‘Not Otherwise Specified’ Rhabdomyosarcomas: 5-Year Survival Rates According to Prognostic Variables (Univariate Analysis)
 5-year EFSa (%)5-year OSb (%)P value
  • HN: head-neck; Ext.: extremities; GU: genitourinary.

  • a

    Five-year EFS = 32.9% for entire cohort of 110 patients.

  • b

    Five-year OS = 45.7% for entire cohort of 110 patients.

Male (n = 59) vs. female (n = 51)34.2 vs. 31.844.7 vs. 43.10.9313
Age 19–30 yrs (n = 69) vs. 31–60 yrs (n = 33) vs. > 60 yrs (n = 8)37.2 vs. 23.6 vs. 31.343.9 vs. 50.3 vs. 48.60.9393
Alveolar (n = 48) vs. nonalveolar (n = 62)25.5 vs. 39.030.1 vs. 58.00.0192
HN (n = 48) vs. Ext. (n = 17) vs. GU (n = 24) vs. other (n = 21)32.6 vs. 20.7 vs. 50.4 vs. 25.946.7 vs. 36.1 vs. 59.9 vs. 38.80.6809
T1 (n = 32) vs. T2 (n = 78)58.4 vs. 21.877.3 vs. 30.9< 0.0001
N0 (n = 68) vs. N1 (n = 42)45.2 vs. 10.156.8 vs. 21.5< 0.0001
Size ≤ 5 cm (n = 33) vs. > 5 cm (n = 77)51.7 vs. 23.875.4 vs. 30.4< 0.0001
IRS Group I (n = 22) vs. Group II (n = 22) vs. Group III (n = 66)49.1 vs. 34.1 vs. 28.081.1 vs. 43.6 vs. 36.90.0103

Pleomorphic RMS

Thirty-seven patients (median age, 50 years; range, 24–83 years) had pleomorphic RMS. Compared with the subset of 143 patients with embryonal, alveolar, or NOS histotype, the pleomorphic RMS subgroup was composed of a greater proportion of males (70% vs. 52%), older patients (84% vs. 32% age > 30 years and 43% vs. 7% age > 60 years), and patients with primary tumor location in the extremities (65% vs. 17%). With regard to TNM staging, 19 cases of pleomorphic RMS were classified as T2b, 6 were classified as N1, and 2 were classified as M1 (for lung metastases). Most patients with pleomorphic RMS (86%) had tumors larger than 5 cm.

Thirty-five patients with pleomorphic RMS were evaluable for treatment and outcome. Complete surgical resection was performed for most patients (74%, compared with 30% for patients with nonpleomorphic RMS). Fourteen patients (including four who received amputations) underwent complete resection at the first surgical attempt; nine underwent complete resection via reexcision within a few weeks of the initial surgery; and three underwent complete resection via delayed surgery, after chemotherapy. Nineteen patients received radiotherapy, and 12 received chemotherapy.

Median follow-up time was 28 months. Five-year EFS and OS were 29.9% and 53.4%, respectively. Twenty-one patients were alive at the time of the analysis (13 after achieving first CR). Twenty patients had recurrent disease (local [n = 11], local with distant metastases [n = 1], and with distant spread [n = 8]). Four of the 11 patients who had local recurrence were alive after achieving a second or third CR.

Surgery was a mainstay of treatment for pleomorphic RMS and was correlated with survival. Five-year EFS was 36.9% for patients who underwent complete resection, whereas no patient with unresectable tumor achieved 5-year EFS; 5-year OS rates for these 2 subgroups were 55.7% and 52%, respectively.

Data regarding tumor response to chemotherapy were available in only two cases (one PR and one lack of response). Distant recurrence occurred in 2 of 12 patients (16%) who received chemotherapy and in 7 of 23 (30%) who did not receive it.

Metastatic RMS

Thirty-one patients (17% of the entire series) had metastatic disease at diagnosis. The median age of patients with metastatic RMS was 21 years (range, 19–61 years), and 84% of patients were younger than age 30 years. Sixteen patients in this group were males. Histologic subtype was embryonal in 14 patients, alveolar in 12, NOS in 3, and pleomorphic in 2. Tumor size and local invasiveness were strongly correlated with metastases at onset: all but one patient with M1 disease had primary tumors classified as T2b.

Metastasis sites were lung (n = 11), bone (n = 6), omentum (n = 5), distant lymph nodes (n = 4), bone marrow (n = 3), soft tissue (n = 3), liver (n = 2), kidney (n = 1), suprarenal gland (n = 1), and brain (n = 1). Six patients had metastases to multiple sites.

Among the 27 patients who were evaluable for treatment outcome, only a 21-year-old male with primary paratesticular embryonal RMS and lung metastases at diagnosis was a long-term survivor with no evidence of disease (alive after achieving first CR, at 137 months from diagnosis). One patient was alive after achieving CR, at 10 months from diagnosis; 1 was alive after achieving a third CR, shortly after a second recurrence; and 2 were alive with disease. The remaining 22 patients died of disease at 2–47 months (median, 14 months) after diagnosis. Five-year OS was 4.3%.


Among 110 adult patients with embryonal, alveolar, or NOS histology, from a series of 180 adults with RMS who were treated at a single institution during a 25-year period and retrospectively reviewed, 5-year EFS was in the 30% range. This rate is lower than rates reported in selected series from collaborative pediatric trials (Table 4)9, 10, 13–15 and closely parallels the results of other published studies of adults with RMS.16–23 However, 5-year OS was 61.5% in the subgroup of 43 patients whose treatment regimens were consistent with current guidelines for standard treatment of pediatric RMS; 5-year OS increased to 72.5% for patients within this subgroup who had embryonal RMS.

Table 4. Improvements in Pediatric Rhabdomyosarcoma Outcomes over Time
 IRS I (1972–1978)IRS II (1978–1984)IRS III (1984–1991)IRS IV (1991–1997)
  • IRS: Intergroup Rhabdomyosarcoma Study; OS: overall survival; RMS: rhabdomyosarcoma.

  • a

    Three-year overall survival.

5-year OS (%)    
 All RMS55637174
 Localized RMS63717986a

Although it certainly is an oversimplification, one might conclude from these findings that there may be no major difference in outcome between adults and children with RMS, provided that patients with the same disease characteristics are considered and that treatment follows the same principles for adults and children. This conclusion disagrees with the view, expressed by some authors, that adult RMS is inherently different from pediatric RMS.16–24 Unsatisfactory treatment results even have raised doubts as to whether chemotherapy should be used at all to treat adults with RMS; Hawkins et al.21 recently concluded that there was no evidence that chemotherapy provided any survival benefit for adult patients with RMS. Nonetheless, we agree with Esnaola et al.22 and Little et al.23 in our opinion that this is not the case. In the current series, aside from the favorable outcome of patients who were treated according to current guidelines for pediatric RMS, the overall rate of response to chemotherapy was 85%. This response rate substantially differs from the rate observed in adults with soft tissue spindle cell sarcomas (response rate < 50%) and clearly falls in the same range as the rate for pediatric small cell sarcomas (i.e., RMS and Ewing family tumors). Chemotherapy appears to have the same activity in adult and pediatric RMS, and when chemotherapy is included in a regimen similar to those used to treat pediatric patients, the outcomes for adults and children with RMS are similar to each other. In the absence of controlled, prospective trials, which, given the extreme rarity of the disease, clearly are unfeasible for adults with RMS, we believe that the findings of the current study are sufficient for recommending that adults be treated according to the same principles that have dramatically improved the prognosis of children with RMS in recent decades.

That said, it remains to be understood why, in all published studies (Table 5), survival of adults with RMS has been decidedly lower than that of children, falling in the range of 20–40%,16–24 just as in the current series taken as a whole. Patients in the current series were treated in the last 25 years; during this time, multiagent systemic treatment has been readily available at a referral European institution for sarcomas that have been involved in pediatric RMS trials. Nonetheless, therapies have been evolving over a quarter of a century, and in a retrospective series such as the current one, one cannot expect to observe the same results as those reported in prospective clinical trials performed in recent years. Still, the observation of similar results in a distinct subgroup in the current series indicates that the question posed by the current study is worth answering. Exploration of the factors that prevent most adult patients (approximately 60% in the current series) from receiving fully adequate treatment is called for. Of course, treatment appropriateness cannot be regarded as an independent variable. Patients with high scores for treatment appropriateness were simply those who were capable of receiving relatively intensive treatment. Having more favorable prognostic indicators at onset probably allowed these patients to receive intensive treatment. However, this may be the critical point—which prognostic factors affect adult series as compared with pediatric series?

Table 5. Literature Review of Adult Rhabdomyosarcoma
Study (institution)No. of patientsStudy periodTreatment (% of patients)ResultsConclusions/comments
  • RT: radiotherapy; OS: overall survival; MSKCC: Memorial Sloan-Kettering Cancer Center; CT: chemotherapy; EFS: event-free survival; RMS: rhabdomyosarcoma.

  • a

    Five-year overall survival showed a clear decrease with increasing age.

  • b

    Five-year event-free survival was approximately 50% among patients younger than age 20 years and < 20% among patients older than age 20 years.

Ariel, 197524 (Park Medical Foundation, New York, NY)1131940–1970RT (46)5 yr OS = 32%
Lloyd et al., 198316 (MSKCC, New York, NY)541950–1978RT (44) CT (37)5 yr OS = 21%Tumor size and resectability are prognostic factors
Seidal et al., 198918 (Sahlgren Hospital, Gothenburg, Sweden)121972–1981CT (8)3 of 12 patients alivePathologic review from the Swedish Cancer Registry
Prestidge and Donaldson, 198924 (Stanford University, Stanford, CA)161967–1987RT (100) CT (81)5 yr OS = 22%Advanced disease in 87% of patients Contains discussion of possible explanations for poor outcomes for adult patients with ‘pediatric’ tumors
Kattan et al., 199319 (Institut Gustave-Roussy, Paris, France)131975–1991CT (92)50% rate of response to CT 8 of 13 patients died of diseaseParatesticular site Unfavorable clinical presentation, disappointing outcome
La Quaglia et al., 199420 (MSKCC, New York, NY)133 adults 157 children1970–1991RT (67) CT (95)5 yr OS = 56%aComparison between adults and children Age is an independent prognostic factor
Hawkins et al., 200121 (MSKCC, New York, NY)841982–1999RT (49) CT (93)5 yr EFS = 35%b No reported data on rate of response to CTAdult RMS behaves similarly to other adult sarcomas No evidence that CT is of any benefit
Esnaola et al., 200122 (Dana-Farber Cancer Institute, Boston, MA)391973–1996RT (69) CT (95)5 yr OS = 31% 82% rate of response to CTMetastatic relapse was the primary cause of failure CT response and survival are correlated
Little et al., 200223 (M. D. Anderson Cancer Center, Houston, TX)821960–1998RT (100) CT (71)5 yr OS = 44%Tumor size is the principal prognostic factor Improved outcome in patients treated with CT (not statistically significant)

On average, clinical presentation in the current series was less favorable than in pediatric series. Table 6 shows the clinical characteristics of patients in the current series along with those of a group of 252 pediatric patients who concomitantly were enrolled in the ICG RMS-88 study. In no way can this be viewed as a formal comparison; it serves only to provide a rough idea of the possible disparities between adults and children with RMS. In the current series, the proportion of patients with alveolar RMS was greater (31% [40% if pleomorphic RMS is excluded] vs. 27%); invasive tumors (65% vs. 44%) and large tumors (73% vs. 50%) were more common; and there was a higher rate of lymph node involvement (31% vs. 15%). These differences may explain in part the less favorable outcome of adult patients overall, and they invite us to treat adults with intensive approaches.22, 23 But are intensive treatments feasible for adult patients?

Table 6. Comparison of the Current Series with the Pediatric Series from the Italian Cooperative Group RMS-88 Protocol
 RMS-88 seriesaCurrent series
  • RMS: rhabdomyosarcoma; NOS: not otherwise specified; IRS: Intergroup Rhabdomyosarcoma Studies.

  • a

    Data from Carli M, Bisoquo A, Guglielmi M, et al. Improved outcome for children with embryonal rhabdomyosarcoma: results of the Italian cooperative study (abstract O87). Med Pediatr Oncol. 2000;35:191.28

No. of patients252180
Study period1988–19951975–2001
Age (yrs)  
Metastatic disease3431
Localized disease218149
 Histology (%)  
 Site of origin (%)  
 TNM classification (%)  
 Tumor size (%)  
  > 5 cm5073
 IRS group (%)  

In principle, an adult might tolerate treatments designed for children to a lesser degree.25 In the current series, we were unable to analyze dose intensity, and it generally is difficult to understand retrospectively a patient's clinical status and fitness for aggressive treatment. Less than half of the patients in the report by Prestidge and Donaldson25 were estimated to have received the full scheduled chemotherapy regimen. Nonetheless, one might point out that it has been demonstrated in several tumors that adults can tolerate dose-intense, multimodal treatments. However, regimens specifically designed for children may not be the most suitable ones for adults; adults may be able to tolerate regimens as intense as those used to treat children, but not necessarily those developed specifically for children. One also might guess that the opinions of the attending medical oncologist may make a difference, at least in part, in terms of what type of treatment is used. We hope that series such as the current one convince medical oncologists that intensive treatment may be worthwhile in adults with RMS.

In the current study, as well as in other adult series, there was a disproportionately high incidence of pleomorphic RMS. Pleomorphic RMS needs to be considered separately from other RMS subtypes. It is very rare in the pediatric population, accounting for less than 1% of all RMS cases in the IRS I–III studies. In contrast, it is common in adults, typically arising in the deep soft tissues of the extremities of patients (predominantly males) older than age 45 years.1 In the past, pleomorphic RMS was diagnosed quite frequently, and later it was regarded as a variant of malignant fibrous histiocytoma. More recently, ultrastructural, immunohistochemical, and molecular techniques have led to refinement of the criteria for diagnosis.26, 27 Pleomorphic RMS is an aggressive neoplasm that probably is closer, both biologically and clinically, to adult, high-grade soft tissue sarcomas than to pediatric RMS. Adult oncologists tend to treat pleomorphic RMS as nonpediatric soft tissue spindle cell sarcomas. The activity of chemotherapy against pleomorphic RMS may be closer to the observed activity against adult sarcomas, and the role of chemotherapy in multimodality treatment is less clear for pleomorphic RMS than for pediatric sarcomas.1 In the current series, data on response to chemotherapy were available in only two cases (one PR and one lack of response). Distant recurrence occurred in 16% of patients who were treated with chemotherapy and in 30% of those who did not receive it. It is noteworthy that of the 23 patients with pleomorphic RMS who did not undergo chemotherapy, 16 did not experience recurrence. This finding would be unexpected in a series of patients with typical, nonpleomorphic RMS. Although pleomorphic RMS is one of the adult soft tissue sarcomas that, being high-grade, appear most likely to respond to adjuvant chemotherapy, response levels such as those found in typical RMS were not observed.3

There is little information to add regarding patients with metastatic disease. Their outcome was exceedingly poor in the current series, although this typically is the case in large pediatric series as well.

In conclusion, the current study confirms that on average, the outcome of adults with RMS appears to be worse than that of children. This likely is due in part to a higher proportion of adults presenting with poor prognosis as compared with children. In addition, pleomorphic RMS is overrepresented in adult series (although how often this diagnosis currently is made by pathologists remains to be determined). Nonetheless, when series are stratified according to known prognostic factors and appropriateness of treatment, prognosis of RMS in adults may be roughly similar to prognosis in pediatric patients. This finding, along with the observation that RMS responds to chemotherapy in adults exactly as it does in children, suggests that RMS in adults is unlikely to be fundamentally different from RMS in children. It is possible that age simply affects the relative size of various prognostic subsets in adults; this holds true within the pediatric population as well, where age is considered an adverse prognostic factor.9 Therefore, we believe that there is no reason to treat adults with RMS according to different guidelines from those that have been established over the years in large pediatric trials. To maximize the likelihood that adult patients are treated according to these principles, protocols that include drug regimens and treatment combinations specifically designed for adults might be worth trying prospectively. Even uncontrolled studies of adults could be useful, provided that they are designed in a way that makes them suitable for confirming the applicability in another age group, through specific treatment solutions, of principles established in large, controlled trials in pediatric populations.