INTRODUCTION — Rhabdomyosarcomas (RMS) are malignant soft tissue tumors that are thought to originate from immature cells and myogenic satellite cells that are destined to form striated skeletal muscle; however, these tumors can arise in locations where skeletal muscle is not typically found (eg, the urinary bladder).
The treatment of RMS has evolved considerably. Cure rates have risen, largely due to the use of combined modality therapy in trials conducted by large international cooperative groups, such as the Intergroup Rhabdomyosarcoma Study Group (IRSG), which is also known as the Soft Tissue Sarcoma Committee of the Children's Oncology Group (COG) [1-3].
This topic review will provide an overview of treatment for RMS. The epidemiology, pathology, molecular pathogenesis, clinical presentation, diagnostic evaluation, and staging of RMS are discussed elsewhere. (See "Rhabdomyosarcoma in childhood and adolescence: Epidemiology, pathology, and molecular pathogenesis" and "Rhabdomyosarcoma in childhood and adolescence: Clinical presentation, diagnostic evaluation, and staging".)
TREATMENT IN CHILDHOOD AND ADOLESCENCE
Overview of risk-adapted therapy — In the past, fewer than 20 percent of patients with rhabdomyosarcoma (RMS) were cured with surgery alone, implying that micrometastatic disease was present at diagnosis in the majority of these patients [4]. In contrast, with the use of modern combined modality therapy, over 70 percent of children with localized RMS can be cured of their disease [3,5].
These improved outcomes are the direct result of multimodality therapeutic protocols that have been developed by large international cooperative groups, such as the Intergroup Rhabdomyosarcoma Study Group (IRSG, or the Soft Tissue Sarcoma Committee of the Children's Oncology Group [COG]). Modern treatment for RMS includes chemotherapy for primary cytoreduction and eradication of both macroscopic and microscopic metastatic disease; surgery, if feasible; and radiation therapy (RT) to control microscopic local residual disease.
The specific treatment regimen depends on the estimated risk of a disease recurrence, which is based upon a variety of clinicopathologic prognostic factors that have been identified through a series of treatment protocols carried out by the IRSG. This is termed risk-adapted therapy. One risk stratification schema is used in the United States (table 1). A different risk stratification system that does not incorporate molecular classification (fusion gene status) is used in Europe [6]. (See "Rhabdomyosarcoma in childhood and adolescence: Epidemiology, pathology, and molecular pathogenesis", section on 'Molecular classification and risk stratification'.)
In this topic, we describe treatment guidelines based upon the established historical definition of "risk group," (table 1), and not the risk groups defined in individual protocols. (See 'Intermediate risk' below and 'High risk' below.)
Overview of IRSG study results — The IRSG, a National Cancer Institute (NCI)-sponsored cooperative group, was formed in 1972 to investigate the biology and treatment of RMS; it was merged into the COG in the year 2000.
Since its inception, six protocols involving over 5000 patients have been completed: IRS-I (which accrued between 1972 and 1978), IRS-II (1978 to 1984), IRS-III (1984 to 1991), IRS-IV Pilot (1987 to 1991), IRS-IV (1991 to 1997), and the D series protocols of IRS-V (1996 to 2005) [7-12]. These studies have been instrumental in identifying prognostic factors, developing risk-based therapies, and improving outcomes for RMS. The group's success is reflected in dramatic increases in the overall five-year survival rates for patients with RMS, from 55 percent in IRS-I to 63 percent in IRS-II to approximately 71 percent in the IRS-III and IRS-IV protocols [12-14].
All of these studies enrolled all stages of disease, but separate protocols were carried out for completely resected localized disease, incompletely resected localized disease, and regionally advanced/distant disease. The most important findings of these studies, which have formed the basis for risk-adapted therapy, can be summarized as follows:
●The extent of surgical resection (ie, the clinical group [CG] classification (table 2)) correlated with outcome in IRS-I, II, and III [7,9,10]. In IRS-III, for example, patients with completely resected RMS (CG I) had a five-year survival rate of over 90 percent, as compared with 80, 70, and 30 percent for those with CG II, III, and IV disease, respectively. For CG III patients (gross residual disease after surgery), neither a higher total dose of RT nor the use of accelerated fractionation RT improved local control or survival [15]. (See "Rhabdomyosarcoma in childhood and adolescence: Clinical presentation, diagnostic evaluation, and staging", section on 'Clinical group'.)
●The tumor, node, metastasis (TNM) staging system (table 3) was prospectively implemented in IRS-III after it was validated using data from IRS-II [10,16,17]. For IRS-IV, TNM stage was used to assign chemotherapy treatment, while CG was used to select patients for RT. In subsequent studies (eg, IRS-V), a combined prognostic stratification scheme was used to assign all therapeutic modalities based upon estimated prognosis (ie, risk-adapted therapy) [16]. (See "Rhabdomyosarcoma in childhood and adolescence: Clinical presentation, diagnostic evaluation, and staging", section on 'TNM system' and "Rhabdomyosarcoma in childhood and adolescence: Clinical presentation, diagnostic evaluation, and staging", section on 'Prognostic stratification scheme for risk-adapted therapy'.)
●The schema used for risk-adapted therapy is outlined in the table (table 1). Options for standard treatment based upon risk group designation are listed in the table (table 4). These include two- or three-drug regimens (vincristine plus dactinomycin [VA] or VA combined with reduced-cumulative-dose cyclophosphamide [VAC]) for patients with low-risk tumors at sites such as the orbit, paratesticular sites, and the head and neck (nonparameningeal) [7,9,10]. For others, a three-drug regimen (VAC with a higher cumulative dose of cyclophosphamide) is recommended. The addition of other individually active agents (eg, doxorubicin, cisplatin, etoposide, ifosfamide, topotecan, or melphalan) has not been shown to improve outcome in any patient subgroup. (See 'Overview' below.)
The following sections provide an overview of standard treatment recommendations for RMS, which are based mainly upon the results of IRS-IV, the D series studies of IRS-V, and all of the COG-era ARST protocols.
An important point is that treatment for RMS is constantly evolving, and if the reader's goal is to obtain information on how to treat a specific patient, he/she should contact a pediatric oncologist who is a member of the COG and who has expertise in soft tissue sarcoma for the latest recommendation. All eligible patients with RMS should be enrolled in one of the active clinical research protocols in any part of the world. In the United States, these are designed by the Soft Tissue Sarcoma Committee of the COG and administered by COG centers according to risk groups (table 1). For available open protocols, a member center should be contacted. Although in the United States pediatric oncologists treat patients up to the age of 20, any patient younger than 50 years old is eligible for these protocols.
Role of surgery — We recommend complete excision for localized disease as long as functional and/or cosmetic results are acceptable. If complete resection is not feasible or if disease involves the orbit, vagina, bladder, or biliary tract, patients are better approached with an initial diagnostic biopsy followed by neoadjuvant chemotherapy and then definitive local therapy with either surgery or RT. For disease at any primary site, palpable or radiologically enlarged (≥1 cm) lymph nodes should be excised and submitted for histologic evaluation, regardless of positron emission tomography (PET) avidity, if this modality is used for staging. (See "Rhabdomyosarcoma in childhood and adolescence: Clinical presentation, diagnostic evaluation, and staging".)
The necessity of routine surgical sampling of nonsuspicious lymph nodes depends on the primary site and will be covered in the sections below.
Head and neck — Primary head and neck tumors are rarely amenable to wide local excision because of proximity to vital structures and cosmetic concerns. This generally limits the role of surgery to initial diagnostic biopsy. However, the availability of highly skilled otolaryngologic and craniofacial reconstructive teams at selected institutions may permit resection of some tumors that would otherwise be considered unresectable [18]. Because of the low incidence of locoregional node involvement (approximately 20 percent, no different from other sites except for the orbit [19]), routine cervical lymph node sampling is unnecessary. This subject is discussed in detail elsewhere. (See "Head and neck sarcomas", section on 'Rhabdomyosarcoma'.)
Paratesticular — Paratesticular tumors should be removed via radical inguinal orchiectomy and resection of the entire spermatic cord. As with testicular germ cell tumors, a transscrotal surgical approach is specifically contraindicated in order to avoid scrotal contamination [20]. (See "Radical inguinal orchiectomy for testicular germ cell tumors".)
All boys 10 years of age or older who have a paratesticular RMS and no imaging evidence of gross lymphatic metastatic disease should undergo routine ipsilateral nerve-sparing retroperitoneal lymph node (RPLN) sampling. The rationale for this recommendation is as follows:
●Overall, approximately 14 percent of boys with paratesticular RMS without radiographic lymph node enlargement will be found to harbor metastases; the risk is much higher (up to 47 percent) among adolescents [21-23].
●In IRS-III, RPLN sampling was recommended, and only 2 of the 11 failures (18 percent) were regional. In contrast, routine RPLN sampling was not used in IRS-IV, and 7 of the total 16 failures (43 percent) were nodal. For patients over the age of 10 who had CG I paratesticular RMS and only received VA, the three-year failure-free survival (FFS, defined as the time from therapy initiation to disease progression, recurrence, or death from any cause) rate was significantly worse than that for younger boys with CG II disease (68 versus 100 percent) who were treated with intensive chemotherapy and RT [22].
●The histopathologic documentation of nodal metastases is clinically important because patients with positive nodes are referred for postoperative RT as well as three-drug (VAC) chemotherapy, while both cyclophosphamide and RT can be withheld in those with negative nodes; FFS in these adequately staged patients is 100 percent.
For younger boys (under the age of 10), surgical excision of radiographically enlarged lymph nodes is warranted, but RPLN sampling is not indicated in the absence of definite radiographic findings.
Other genitourinary tumors — The functional morbidity of initial surgery for other genitourinary tract tumors has prompted the avoidance of upfront surgery in favor of multimodality approaches [24,25]. Vulvar, vaginal, uterine, bladder, and prostatic RMS usually respond well enough to induction chemotherapy to render them locally resectable, often with negative resection margins for tumor involvement [26-29].
One possible exception is a tumor arising in the dome of the bladder, for which partial cystectomy may be performed before chemotherapy [30,31]. The benefit and timing of RT in this setting are controversial [32]. (See 'Radiation therapy' below.)
Extremity and trunk — Despite advances in therapy, outcome for children with an extremity RMS is inferior to that of more favorable sites [7,33]. Initial resection of a localized extremity RMS should be attempted as long as limb function is not compromised; outcomes are significantly worse if gross residual disease remains [34,35]. In all cases, patients with extremity RMS should undergo sentinel lymph node biopsy if clinically node negative [36,37]. If the sentinel node(s) is/are positive by histopathologic evaluation, assessment of more-proximal lymph node levels in the chain of lymphatic drainage should be done with additional imaging. Routine ilioinguinal lymphadenectomy is also recommended for perineal or anal RMS [38].
Attempted resection is also appropriate for truncal tumors [39]. Similar to extremity tumors, tumors involving the chest wall appear to do worse after wide surgical excision if the margins are not negative for tumor [21,40]. In all of these cases, re-excision should be considered for questionable or known positive margins.
Response to therapy and second-look surgery — Selected patients may benefit from second-look surgery after upfront chemotherapy to evaluate for residual disease. A multidisciplinary discussion among the surgeon, radiation oncologist, pediatric oncologist, and the parents should take place after chemotherapy in all patients with CG III disease (table 2) regarding the risks and potential benefits involved in a second-look surgery. The main benefit of second-look surgery in this setting is to identify patients who may benefit from reduced-dose RT.
For patients with a clinical (radiographically) complete response to induction chemotherapy, second-look surgery has been studied as a means of documenting the histopathologic response in an attempt to improve outcomes. However, a survival benefit from this approach has not been proven. A lack of correlation between radiographic tumor response to induction therapy and overall prognosis was shown in the IRS-IV trial, in which CG III participants who completed all protocol therapy without developing progressive disease were assessed for response (complete response, partial response [≥50 percent decrease], or no response [<50 percent decrease and <25 percent increase]) by radiographic measurement [41]. When categorized as to the best response to therapy, five-year FFS was similar for participants achieving a clinical complete response and for those whose best response was a partial response or no response (80 versus 78 percent). Among the 17 participants with a best response of partial/no response who had second-look surgical procedures, 8 (50 percent) of 16 with available pathology reports had residual viable tumors, and resection did not improve outcomes (refer below).
However, it should be noted that in other studies, such as the Cooperative Weichteilsarkom Studiengruppe (CWS) RMS trials, partial response (>33 percent reduction in tumor size) or an objective response of a lesser magnitude (<33 percent reduction in tumor size) after three cycles of chemotherapy has been correlated with better outcomes [42].
Additional data examining the value of second-look surgery come from the International Society of Paediatric Oncology (SIOP) Malignant Mesenchymal Tumors study, in which 237 patients with nonparameningeal RMS underwent initial biopsy or limited surgery, followed by three to six cycles of chemotherapy and a second-look surgical procedure [43]. RT was withheld in patients with a histopathologic complete response. Unfortunately, 50 percent of patients ultimately had a local recurrence, regardless of whether they underwent second-look surgery or were followed clinically.
This approach was also evaluated using data from the IRS-III study, in which patients with CG III disease (table 2) underwent induction chemotherapy and local RT; some patients underwent second-look surgery, which was delayed until 20 weeks after therapy initiation, when possible [7]. The goal was to render the patients tumor free by week 20. The following results were noted in a subset of 111 patients with pelvic tumors:
●Twenty-seven patients had a clinical complete response, 55 patients had a partial response, and there were 29 patients with minor responses (objective tumor regression of less than 50 percent of baseline tumor size).
●Among the 55 patients who had a partial response undergoing second-look surgery, 14 (25 percent) were in fact histopathologic complete responses and an additional 12 (22 percent) were rendered tumor free by surgery.
●Five of 29 patients with a minor response were found to have histopathologic complete responses, while seven others were converted to a histopathologic complete response as a result of surgery.
●Despite these benefits, the lack of randomized studies precludes any conclusions as to the survival benefit of second-look surgery to resect residual disease.
The survival impact of a second-look surgical procedure during treatment was also addressed in an analysis of a subset of patients who participated in IRS-IV and who underwent a second-look procedure before the completion of chemotherapy [44]. The initial protocol had stipulated that a second-look procedure be considered for patients who still had a persistent tumor mass at week 46 or 47, approximately four weeks after completing treatment. Despite this, 79 children had a second-look procedure prior to completing chemotherapy. Overall, 13 of the 14 patients with a clinical complete response had no viable tumor, while 41 percent of those without a complete response had no viable tumor. Five-year FFS rates were significantly higher in patients without viable tumor when compared with those who had viable tumor (81 versus 53 percent, hazard ratio [HR] 0.38, 95% CI 0.15-0.99); overall survival did not reach the level of statistical significance (81 versus 67 percent, HR for death 0.42, 95% CI 0.12-1.41).
In another COG study of 369 patients with clinical group III RMS (table 2) who either received delayed primary excision (on protocol D9803) or did not receive delay primary excision (on protocol ARST0531), survival was similar between the two groups [25].
Based upon these reports, there is no clear survival advantage to second-look surgical procedures when they are used to document the response to chemotherapy in children with RMS, and second-look surgery should only be used selectively.
On the other hand, a reasonable indication for second-look surgery is to document the histopathologic response to chemotherapy in order to allow a reduction in the prescribed radiation dose in patients with CG III disease (table 2). In the D9802 protocol, doses as low as 36 Gy (in those with a complete response [complete resection with negative margins]) and 41.4 Gy (in those with positive margins) were employed with equal local control and survival compared with previous protocols for patients who otherwise would have received 50.4 Gy [45].
Metastatic disease — In contrast to other pediatric bone and soft tissue sarcomas, where metastasectomy is often considered, particularly for limited pulmonary metastases, the role of such surgery for patients with metastatic RMS is unclear (see "Surgical treatment and other localized therapy for metastatic soft tissue sarcoma"):
●In a series of 152 children with a variety of sarcomas undergoing metastasectomy, those with RMS had a uniformly poor outcome [46].
●On the other hand, at least some data suggest that resection of metastatic disease may benefit some patients. In a report of 88 patients with metastatic RMS treated over three consecutive European cooperative group protocols, patients who underwent aggressive attempts at local control (surgery plus radiation) had a longer overall survival compared with those who were treated with surgery or radiation only (44.3, 18.8, and 16.1 percent, respectively, p<0.006) [47].
The decision to pursue metastasectomy must be individualized and made on a case-by-case basis.
Radiation therapy — We recommend RT to enhance local control in all patients with RMS except those who have embryonal/fusion-negative CG I tumors. Although the benefit of RT has yet to be determined, treatment standards in North America also recommend the addition of RT to all patients with alveolar histology and CG I disease, regardless of fusion status.
General issues — RT is a major treatment modality for RMS, particularly for achieving local control in patients with residual microscopic or gross disease following surgery and chemotherapy (ie, CG II or III disease (table 2)). Computerized three-dimensional treatment planning and conformal delivery systems should be applied whenever available to minimize radiation exposure of surrounding normal tissues [48].
Where available, proton beam RT appears to represent a safe and effective alternative radiation modality to photon beam irradiation [49]. Because it can potentially eliminate the exit dose to normal tissue, proton beam irradiation may provide a significant reduction in late effects with orbit [50] and parameningeal tumors [51], and it is permitted in RMS trials.
Timing and treatment volume — Based upon prior cooperative group studies, when it is administered, RT is generally given after four cycles of chemotherapy. Historically, for patients with parameningeal tumors and intracranial extension or cranial nerve palsies, radiation was initiated emergently on day 0. However, data suggest that there is no benefit for early RT when intensive chemotherapy is used and that the functional and survival outcomes are similar if RT is delivered after four cycles of chemotherapy [52-54]. In one COG protocol (ARST0531), RT was delivered at week four using lower cumulative cyclophosphamide doses. Local control rates were significantly lower with this approach compared with a previous COG protocol (D9803) when RT was delivered at week 13 with higher cyclophosphamide doses [55]. Many centers, including ours, recommend RT after four cycles of chemotherapy in all patients with parameningeal tumors unless there is intracranial extension or cranial palsies. However, there is a need for further evidence for optimal timing of RT in patients treated with contemporary chemotherapy regimens that are less intense. In the open COG ARST1431 intermediate-risk trial, the only indication for emergency RT (ie, starting on day 0) is vision loss and spinal cord compression.
Treatment volume is determined by tumor extent at diagnosis, including involved nodes, and should include at least a 1 cm margin of normal tissue. RT is typically given over four to six weeks in once-daily fractions of 1.8 to 2 Gy. Although early guidelines recommended doses as high as 55 to 60 Gy for control of the primary, clinical practice has evolved over time, with the underlying strategy of reduced RT doses to minimize posttreatment complications. However, even with reduced doses, the decision to pursue RT in very young children (under the age of three) is difficult, as there are minimal data regarding the survival benefits [56] and potential long-term effects of RT to various locations in this age group [57]. Treatment must be individualized in this situation.
Site-specific issues — The following sections discuss specific RT recommendations for patients with RMS, which are largely based upon results from IRSG studies.
Nonparameningeal tumors
Clinical group I — Adjuvant RT is unnecessary for patients with CG I embryonal/fusion-negative (any site) RMS (table 2), but it is recommended (RT dose 36 Gy) for all patients with CG I alveolar or fusion-positive RMS [9,10,58]. This recommendation is based upon a report of 439 patients with CG I disease treated in IRS-I, II, and III [58]. Pretreatment factors associated with an inferior outcome included size >5 cm, sites other than the genitourinary tract, and alveolar histology. Patients with alveolar histology who received adjuvant RT had significantly better FFS and overall survival than those who did not.
The recommendation that all CG I alveolar RMS receive RT has been called into question by a later analysis of 71 patients with CG I alveolar RMS who were treated in the IRS-III and IV studies [59]. RT was assigned and not randomized, and 30 patients received chemotherapy plus RT, while 41 were treated with chemotherapy alone. Among the 54 patients with stage 1 (favorable site, any size, any N status, M0) or stage 2 (unfavorable site, size ≤5 cm, N0 or Nx, M0 (table 3)) CG I disease, eight-year event-free survival (EFS) was similar with and without RT (90 versus 88 percent), as was local control (100 versus 92 percent). However, the small number of patients with stage 3 (unfavorable site, >5 cm, or ≤5 cm and N1 (table 3)) disease did not have a favorable outcome in the absence of RT (eight-year EFS was 84 percent with 100 percent local control in the 13 patients given RT; eight-year EFS was only 25 percent with 50 percent local control in the four patients treated without RT).
These numbers are very small, and larger studies are needed before it can be concluded that RT is not needed in patients with alveolar histology and CG I disease. Treatment standards in North America recommend the addition of RT to all patients with CG I tumors that are fusion positive (forkhead box O1 [FOXO1] rearranged) or are fusion negative with alveolar histology, but not for CG I tumors that are fusion negative or embryonal histology (if fusion testing was not performed).
Clinical group II — RT is recommended for all patients with residual microscopic disease after surgery (CG II (table 2)). In this group, RT doses from 41.4 to 45 Gy are sufficient to achieve local control [60]. In the IRS-III and IV studies, this approach resulted in five-year FFS rates of 75 and 87 percent, respectively [61].
The available data suggest that lower RT doses (36 Gy) are adequate in conjunction with systemic chemotherapy for patients with microscopic residual disease in the absence of nodal involvement [12,13,62,63]. For all histologies, the total dose is increased to 41.4 Gy if there is evidence of nodal involvement, and the field includes the involved lymph nodes.
Clinical group III — RT is recommended for all patients with CG III disease. The benefit of RT is supported by a pooled analysis of patients with bladder/prostate primaries treated in United States and European cooperative protocols, in which patients who did not receive RT for local control had an inferior EFS, although overall survival was not adversely affected [64].
It may be possible to omit RT in certain subsets of patients, but this is not the standard of care in the United States yet and needs to be appropriately studied. As an example, for patients with localized vaginal disease, the last two United States trials, D9602 [12] and ARST0331 [65], attempted to omit RT based on response to chemotherapy. However, this approach led to an unacceptably increased rate of local failure [66]. Patients with group IIA or III N0 disease who did not receive RT had a 46 percent incidence of local recurrence when treated with VAC/VA according to the ARST0331 protocol, compared with 14 percent when treated with VAC high-dose cyclophosphamide according to the D9602 trial [66]. Thus, these higher local failure rates may be attributed to the lower doses of cyclophosphamide; a similar pattern of higher local failure rates has been observed among patients with intermediate-risk disease who received lower total cumulative doses of cyclophosphamide (table 4) [55]. Therefore, until further information becomes available, for any CG III vaginal RMS patient in whom RT is not feasible or is not going to be administered, we recommend higher total cumulative doses of cyclophosphamide, as administered in D9602 [12].
Patients with gross residual disease (CG III (table 2)) require higher RT doses than those that are used for CG II disease, generally 45 to 50.4 Gy. In IRS-III, the five-year FFS in this group was 70 percent and was better for those with node-negative as compared with node-positive disease (75 versus 45 percent, respectively) [67].
In some United States RMS studies, a dose of 50.4 Gy has been recommended for all sites except the orbit. Later studies support the use of even lower doses in children with CG III disease who have only microscopic disease remaining after induction chemotherapy, as determined by second-look surgery. As an example, in the D9802 protocol, doses as low as 36 Gy (for patients with a complete response after second-look surgery) and 41.4 Gy (for patients with positive margins after second-look surgery) were employed with equal local control and survival compared with previous protocols [45]. This approach should be limited to children undergoing second-look surgery. (See 'Response to therapy and second-look surgery' above.)
Orbital tumors — The optimal RT dose for orbital CG III tumors is unclear, and this is a controversial area. The following data are available:
●The five-year cumulative incidence of local failure for patients in the D9602 protocol (n = 77) with embryonal CG III orbital tumors treated with 45 Gy without cyclophosphamide (VA only) was 14 percent [13]. This is higher than the five-year cumulative local failure rate of 2 percent for similar patients treated in IRS-IV (n = 49), who received 50.4 to 59.4 Gy and alkylating-agent-containing chemotherapy. However, it is similar to the 16 percent cumulative local failure rate seen in CG III orbital tumors treated in IRS-III (n = 71), who received 41.4 to 50.4 Gy RT (depending on patient age and tumor size) without an alkylating agent.
●In the low-risk RMS COG protocol (ARST0331, n = 62), these patients were treated with four cycles of VAC and 12 additional cycles of VA with 45 Gy RT for local control [68]. The cumulative five-year local failure rate was 13 percent overall, but it was 0 percent in the 15 patients who achieved a complete response after 12 weeks of chemotherapy.
Taken together, these data suggest that reducing the RT dose to 45 Gy for orbital CG III tumors may not result in inferior local control for children when cyclophosphamide is part of the regimen and there is a good response to upfront chemotherapy. However, the number of failures in these comparisons is very small, and there is no legitimate statistical comparison. The result seen in the IRS-IV trial may in fact be the anomaly because it has the smallest number of patients treated among the four studies. It is difficult to make sound conclusions as to whether one RT dose is superior to the other based upon these limited data. In our view, any of these treatment approaches is appropriate for CG III orbital tumors.
Parameningeal sites — For patients with residual disease in parameningeal sites with or without intracranial extension, improvements in RT field definition and the routine use of higher RT doses (up to 55 Gy) have led to dramatic improvements in survival [54,69-73]. RT (minimum dose 50.4 Gy) should be administered to the primary site, the adjacent meninges, and the region of intracranial extension. Because it can potentially eliminate the exit dose to normal tissue, proton beam RT can provide superior sparing of normal tissue and a significant reduction in late effects [69]. (See "Radiation therapy techniques in cancer treatment", section on 'Particle therapy'.)
If there are tumor cells in the cerebrospinal fluid, suggesting diffuse meningeal disease, craniospinal RT is indicated. If there are multiple parenchymal brain metastases and there are no tumor cells in the cerebrospinal fluid, whole-brain RT should be administered.
Role of brachytherapy — For children with small, critically located tumors (head and neck, bladder, prostate, vulva/vagina), intracavitary or interstitial implants (brachytherapy) may be considered in an attempt to deliver RT to a restricted tissue volume with preservation of anatomy and organ function, and less scatter to adjacent structures [74,75]. The use of brachytherapy is limited since it is available in very few centers in the world.
Role of radiation therapy in metastatic disease — For patients with overt metastatic disease, RT can be used to control both the primary and metastatic sites. The total dose is 50.4 Gy for all sites (except the orbit, where the total dose is 45 Gy). The role of whole-lung RT (generally to 14.4 Gy) for patients with overt pulmonary metastases is unclear; some protocols recommend it, particularly for those undergoing surgical resection, given the radiosensitivity of RMS.
Chemotherapy — We recommend chemotherapy in addition to local therapy for all patients with RMS. Chemotherapy is used for primary tumor cytoreduction and for eradication of macroscopic and microscopic metastatic disease. Chemotherapy may be administered as induction therapy or following surgery. The choice of regimen depends on the estimated risk of a disease recurrence (table 4).
Overview — Prior to the advent of combined modality therapy, local therapy alone for RMS resulted in survival rates of <20 percent, presumably due to the presence of subclinical micrometastatic disease in most patients. The incorporation of systemic chemotherapy into treatment protocols has increased survival rates in patients with apparently localized disease to approximately 60 to 90 percent. However, outcomes are worse in certain subsets, including children under the age of 1 year or over age 10 and those with alveolar histology and fusion-positive tumors, unfavorable tumor sites, tumor size larger than 5 cm, and higher CG risk strata (table 2) [76]. (See "Rhabdomyosarcoma in childhood and adolescence: Clinical presentation, diagnostic evaluation, and staging", section on 'Prognostic stratification scheme for risk-adapted therapy'.)
With varying doses and schedules, VA or VAC has been considered the gold-standard regimen for RMS in North America, against which other, more or less aggressive combinations have been compared:
●In IRSG and COG trials, the addition of many individually active agents (eg, doxorubicin, cisplatin, etoposide, ifosfamide, topotecan, and melphalan) did not improve outcomes compared with VAC in any subgroup [14,32,77-82]. In a European trial conducted by the EpSSG in patients with high-risk nonmetastatic RMS, the addition of dose-intensified doxorubicin to ifosfamide, vincristine, and dactinomycin (IVA) also did not improve EFS and increased toxicity [83].
●Due to the absence of a survival improvement with an increasing cumulative dose and intensity of cyclophosphamide in IRS-III, IRS-IV, and D9803, the COG Soft Tissue Sarcoma Committee decreased the cyclophosphamide dose (from 2200 to 1200 mg/m2 per cycle and from 30.8 to 16.8 g/m2 cumulative dose) in the ARST series (eg, ARST0331, ARST0431, and ARST0531) [84,85].
●Irinotecan plus VAC (VAC/VI; with a cumulative cyclophosphamide dose of 8.4 g/m2) was compared with VAC alone in the ARST0531 trial for patients with intermediate-risk RMS (defined as patients with nonmetastatic alveolar tumors and patients with stage 2 or 3, group III embryonal tumors) [84]. Similar overall survival and EFS rates were observed between the two arms, while a lower incidence of hematologic toxicity was seen in the VAC/VI arm. Based on these results, the COG Soft Tissue Sarcoma Committee adopted VAC/VI as the standard-of-care backbone chemotherapeutic regimen for the intermediate-risk trial ARST1431; however, final analysis of the ARST0531 trial has confirmed a higher rate of local recurrence and a lower EFS for group III patients when compared with those treated in the D9803 study, attributed possibly to the lower cyclophosphamide doses [55]. Based on these results, the COG Soft Tissue Sarcoma Committee has incorporated six cycles of maintenance therapy with oral cyclophosphamide and vinorelbine for patients who are enrolled in the ARST1431 study. Although the approach of maintenance chemotherapy is practiced in Europe, off protocol, we do not consider this a standard approach. (See 'Role of maintenance chemotherapy' below.)
●In contrast, the elimination of cyclophosphamide from adjuvant chemotherapy (ie, VA as compared with VAC (table 4)) did not compromise outcome in low-risk, subset A disease (table 1) [12].
Importantly, all of the previous trials were conducted when risk-based algorithms to select treatment for RMS combined histologic classification (embryonal versus alveolar) with presurgical stage and postsurgical CG. However, histologic classification is not always straightforward. Although molecular detection of the chimeric transcription factors PAX3-FOXO1 or PAX7-FOXO1, which are found exclusively in alveolar RMS, was initially thought to provide an objective basis for distinguishing between the two major forms of the disease (alveolar versus embryonal), up to 45 percent of RMS with alveolar morphology lack FOXO1 rearrangement or another unique gene fusion. (See "Rhabdomyosarcoma in childhood and adolescence: Epidemiology, pathology, and molecular pathogenesis", section on 'Alveolar RMS and chromosome translocations'.)
Data suggest that these fusion-negative alveolar tumors behave more like embryonal RMS (with a more favorable outcome overall) than fusion-positive alveolar RMS [86,87]. In the United States, risk stratification schema (and protocols, including ARST1431) have evolved away from using histology alone to classify RMS, and this has been replaced by an assessment of the presence or absence of a characteristic fusion gene by fluorescence in situ hybridization (FISH) or reverse transcription polymerase chain reaction (RT-PCR (table 1)). (See "Rhabdomyosarcoma in childhood and adolescence: Epidemiology, pathology, and molecular pathogenesis", section on 'Molecular classification and risk stratification'.)
In the United States, clinical research protocols for patients with RMS are designed by the Soft Tissue Sarcoma Committee of the COG and administered by group centers according to risk groups. The completed sixth-generation protocols are described below. For availability of any open protocols, the reader should contact a COG center. Any patient younger than 50 years old is eligible for these protocols. For patients diagnosed in other parts of the world, cooperative group protocols are available through organizations such as the SIOP, the CWS, and the European paediatric Soft tissue sarcoma Study Group (EpSSG).
Treatment strategies according to risk group
Low risk — Based upon data from IRS-III to IV and D9602 studies, patients with low-risk disease have an estimated three-year FFS rate of >85 percent [32]. Standard treatment recommendations for patients with low-risk RMS, as determined by the most cooperative group protocols D9602 and ARST0331 [12,13,65], are outlined in the table (table 4).
The data supporting these recommendations are summarized below.
In the ARST0331 protocol, low-risk patients with RMS were divided into two groups [65]:
●Subset A included patients with embryonal tumors and:
•Stage 1 or 2, CG I or II disease
•Stage 1, CG III, orbit disease
●Subset B included other patients with embryonal tumors and:
•Stage 1, CG III, non-orbit disease
•Stage 3, CG I or II disease
In the IRS-III and IV trials, excellent survival outcomes were achieved for both of these groups of patients using VA or VAC chemotherapy plus or minus RT (five-year FFS rates 78 to 89 percent and overall survival >90 percent) [7,8].
The D9602 trial attempted to decrease toxicity by reducing RT doses and eliminating cyclophosphamide for the lowest-risk patients [12]. Patients in subset A (refer above) received VA only, while those in subset B received VAC (per-cycle and cumulative cyclophosphamide doses of 2200 mg/m2 and 28.6 g/m2, respectively). Five-year FFS rates for those in subset A were 89 percent, and they were 85 percent for those in subset B. Five-year overall survival rates were >90 percent for all subsets.
ARST0331 studied lower cyclophosphamide doses (1200 mg/m2 per dose) in both subsets. Compared with previous protocols, the total cumulative dose of cyclophosphamide was limited to 4.8 g/m2 and was given with vincristine plus dactinomycin and RT to areas of residual tumor to patients in each of the two subsets. The goal was to improve outcome without significant acute or long-term toxicity for the majority of subset A patients and to maintain the excellent outcome and reduce acute and long-term toxicity for patients in subset B. Treatment duration was 22 weeks for patients in subset A and 46 weeks for those in subsets B. Results were published separately for subset A and subset B:
●Results for patients included in subset A showed that at a median follow-up of 4.3 years, shorter-duration therapy that included lower-dose cyclophosphamide and RT did not compromise outcomes compared with historical cohorts [65]. The estimated three-year FFS and overall survival rates were 89 and 98 percent, respectively. These values compared favorably with outcomes for similar patients treated on the D9602 protocol.
●The results for subset B showed inferior three-year 70 percent FFS rates overall, resulting from a reduced cyclophosphamide dose and elimination of RT in a subset of patients with vaginal primary tumors [88].
In our view, for patients with low-risk (excellent prognosis, subset A) disease, VA only (as per D9602) or four short cycles of VAC followed by four cycles of VA (per ARST0331) are both accepted standard options. Notably, a study comparing treatment costs between the two regimens showed that the regimen used in ARST0331 (four short cycles of VAC followed by four cycles of VA) is on average less expensive per patient [89]. Societal costs could not be included in the analysis.
For patients with subset B disease, higher doses of cyclophosphamide using the D9602 regimen are the standard of care (table 4).
COG initiated a protocol ARST2032 and created a very low-risk group consisting of CG I stage 1 patients without any tumor MYOD1 and TP53 mutations [90]. They will receive therapy for 24 weeks using VA.
Intermediate risk — There is no consensus as to the standard of care for intermediate-risk disease in the United States. The best survival estimates for this group are reported using IRS-III, IRS-IV [15,16], and D9803 protocols with higher cyclophosphamide doses, which are also associated with more early and late side effects.
Therapy for intermediate-risk disease was addressed in the ARST0531 protocol, which randomly assigned intermediate-risk patients to standard VAC chemotherapy or VAC/VI [84]. The cyclophosphamide dose in this trial was 1200 mg/m2. As a study question (whether local control can be improved), RT started early (at week 4 of therapy) for all patients, concurrent with irinotecan as a radiosensitizing agent in the VAC/VI experimental arm. Early results showed that, compared with VAC alone, patients treated with VAC/VI had similar overall survival and EFS, less hematologic toxicity (anemia, thrombocytopenia), and a lower cumulative cyclophosphamide dose, but more diarrhea [84]. Based on this observation, the COG Soft Tissue Sarcoma Committee chose this combination as the standard arm in the ongoing ARST1431 protocol for patients with intermediate-risk disease.
However, in the final analyses, compared with patients treated with the D9803 protocol (which used higher cyclophosphamide doses), increased local failure rates were observed for group III embryonal patients (cumulative incidence 27.9 versus 19.4 percent, p = 0.03) [55]. This translated into worse EFS and overall survival for all patients in multivariable analyses for patients treated in the ARST0531 protocol. These findings, as well as the data on maintenance treatment from the EpSSG, prompted COG to recommend the addition of six months of oral cyclophosphamide and intravenous vinorelbine as maintenance therapy for the intermediate-risk protocol ARST1431. Although the approach of maintenance chemotherapy is practiced in Europe, off protocol, we do not consider this a standard approach. (See 'Role of maintenance chemotherapy' below.)
While there are no randomized studies comparing the D9803 schedule VAC (using 2200 mg/m2 cyclophosphamide per dose) with the regimens that use lower cyclophosphamide doses (as were used in ARST0531 regimen A) or VAC/VI (as in ARST0531 regimen B) [84], given the inferior results reported in the ARST0531 study, we recommend higher total cumulative doses of cyclophosphamide (as in D9803) for standard therapy of intermediate-risk disease, off protocol. This recommendation includes patients with metastatic (CG IV, stage 4) embryonal tumors who are younger than 10 years of age (table 1), even though they were not included in ARST0531.
A comprehensive discussion about the complex issue of cyclophosphamide dosing in RMS is available [91].
Role of maintenance chemotherapy — Outcomes for children with intermediate-risk disease have remained virtually unchanged over the past two decades despite intensification of alkylating agents and addition of new therapies such as topotecan [14,92]. In an attempt to improve the outcome of these patients, the EpSSG conducted a randomized study in patients with high-risk disease (which is comparable to the intermediate-risk group as defined by the COG). Patients aged 6 months to 21 years with node-negative alveolar or incompletely resected (group II or III) embryonal RMS arising in an unfavorable site (head and neck, pelvis) or with lymph node-positive disease, and who had no radiologic evidence of disease at the end of treatment (typically, six to nine cycles of intense chemotherapy with ifosfamide, vincristine, dactinomycin with or without doxorubicin, surgery, and/or RT) were randomized 1:1 to six months of maintenance therapy using weekly intravenous vinorelbine (25 mg/m2 on days 1, 8, and 15 every 28 days) plus oral cyclophosphamide (25 mg/m2 orally, daily, continuously) or no maintenance therapy [93,94]. Maintenance therapy was associated with a borderline improvement in three-year disease-free survival (the primary endpoint, 77.6 versus 69.8 percent, HR 0.68, 95% CI 0.45-1.02, p = 0.06) and better three-year overall survival (86.5 versus 73.7 percent, HR 0.52, 95% CI 0.32-0.86). Toxicity was predominantly low grade and hematologic. Although 91 percent completed the maintenance treatment, 79 required at least one cycle modification. The authors concluded that this therapy established the new standard of care for this group of patients in Europe.
Several differences exist between this treatment schema and the one used by the COG for American children, including a longer duration of therapy (14 cycles with a higher cyclophosphamide dose) and inclusion of all in the outcome analysis regardless of remission status at the end of a prespecified time point. Thus, in our view, although maintenance therapy is a reasonable approach for children treated according to the EpSSG guidelines, these results cannot be extrapolated or incorporated into the standard therapy regimen for patients treated off protocol according to COG protocols using high cumulative doses of cyclophosphamide. As noted above, maintenance chemotherapy is recommended for patients receiving therapy within the open intermediate-risk protocol ARST1431. (See 'Intermediate risk' above.)
Alveolar histology with negative FOXO1 fusion — Patients with alveolar histology and negative FOXO1 fusion status have favorable survival outcomes that are similar to those with embryonal histology and negative FOXO1 fusion status. (See 'Overview' above.)
Most of these patients will belong to the low- or intermediate-risk category. However, the decision to determine standard chemotherapy in this group should be based on previous best evidence, which has always been using intermediate-risk therapy. Therefore, we treat with higher total cumulative doses of cyclophosphamide (as in the D9803 protocol). They are eligible to be enrolled in ARST1431 and can receive VAC/VI with or without temsirolimus if they have intermediate disease. Under the same protocol, low-risk patients are treated with VAC/VA therapy according to ARST0331 Regimen A.
High risk — High-risk disease includes all patients with metastatic disease with alveolar histology or fusion-positive tumors, and patients with metastatic embryonal or fusion-negative tumors who are older than 10 years of age at diagnosis. Overall five-year EFS for a patient with high-risk disease ranges between approximately 60 and 70 percent with standard therapy [95].
Importantly, not all patients presenting with metastatic disease do poorly with standard therapy, and there are factors other than fusion status that are predictive of outcome. A large pooled dataset using cooperative trial results from Europe and the United States showed that the following variables are poor risk factors and predict survival accurately (Oberlin criteria): age <1 or ≥10 at diagnosis, unfavorable site (limbs or other), bone or bone marrow involvement, and three or more metastatic sites [96].
To improve on the historically poor results in this subgroup, the COG ARST0431 protocol evaluated intensive multiagent therapy with vincristine plus irinotecan and included dose-compressed cycles of VDC (vincristine plus doxorubicin and cyclophosphamide) alternating with ifosfamide/etoposide (IE; ie, treatment cycles were given every two rather than three weeks through use of hematopoietic growth factor support) in 109 patients with metastatic disease at presentation [85]. There were 20 patients with metastatic (CG IV, stage 4) embryonal tumors who were younger than 10 years of age (intermediate risk, historically) included in the cohort. RT to the primary site was administered at weeks 20 to 25 but was permitted at weeks 1 to 6 for intracranial or paraspinal extension and at weeks 47 to 52 for extensive metastatic disease. At a median follow-up of 3.8 years, three-year EFS was 38 percent, and overall survival was 56 percent. Patients with fewer than two Oberlin risk factors [96] had a three-year EFS of 69 percent, a result that compares favorably with a historically treated cohort of patients with high-risk disease in the Oberlin cohort (44 percent EFS) [96]. In contrast, high-risk patients with two or more Oberlin risk factors did not benefit from this approach and had a three-year EFS of 20 percent. For the 20 patients with metastatic (CG IV, stage 4) embryonal tumors who were younger than 10 years of age, three-year EFS was 60 percent, which is similar to the three-year EFS of 59 percent for those treated on D9803 with VAC. (See 'Intermediate risk' above.)
Patients with high-risk disease, particularly if they have two or more Oberlin factors, have a very poor prognosis. There is clearly a need for novel treatment strategies.
We prefer that eligible patients with high-risk disease be enrolled in a clinical trial testing new strategies. We do not recommend treatment according to ARST0431 as an option for this group of patients since the published results do not suggest any improvement over historical cohorts and the regimen is longer, has more side effects, and is more expensive [85]. If a clinical trial is not available, we recommend VAC alone according to regimen A of the ARST0531 protocol in this subset, using lower doses of cyclophosphamide. We expect that a very small number of patients, if any, will be cured. However, they will receive an outpatient regimen with fewer side effects compared with higher intensity regimens; they are likely to have a clinical and imaging response, and they are likely to have palliation of symptoms and a better quality of life.
High-dose chemotherapy — In view of the poor prognosis associated with high-risk disease, high-dose chemotherapy with or without total body irradiation and autologous hematopoietic stem cell transplantation has been studied as an initial treatment strategy. In uncontrolled series that generally involve small numbers of patients, the survival rates using this approach for previously untreated children with metastatic RMS/undifferentiated sarcoma range between 19 to 44 percent [97-100]. However, it is unclear if these results are better than those that can be achieved with standard-dose chemotherapy in RMS since there are no randomized trials directly comparing both approaches. In the only randomized trial exploring the value of high-dose chemotherapy in this setting, the control group received oral maintenance therapy, which was not given to those undergoing high-dose therapy, and only 86 of the 295 registered patients were reported; 162 never received the protocol therapy [101].
Two meta-analyses concluded that there was no significant survival advantage to high-dose chemotherapy/stem cell rescue for patients with advanced RMS [102,103], and a greater risk of grade 3 or 4 toxicity compared with conventional chemotherapy [102].
Thus, the benefits of high-dose chemotherapy are unproven, and this modality should not be pursued unless within the context of a clinical trial.
Surveillance imaging for recurrent disease — The optimal frequency of surveillance imaging to detect recurrent disease is not established. Based on observational data suggesting a similar survival benefit for surveillance imaging versus patient-reported symptoms [104,105], European cooperative groups guidelines recommend a lower number of surveillance imaging (six studies every four months) ending two years from end of therapy in all patients [106]. There is no similar official guideline in the United States and patients are usually imaged per COG treatment protocols over four years with a total of 12 imaging time points.
Observational studies have evaluated the impact of surveillance imaging on detecting recurrences and overall survival. A retrospective analysis of 199 patients with RMS and localized relapses who were enrolled in several European cooperative protocols showed similar overall survival regardless of whether the recurrence was detected by surveillance imaging or a symptom (50 versus 46 percent) [104]. Similar results were seen in another retrospective study conducted in the United States of 127 patients with localized and metastatic disease at relapse [105].
Recurrent disease — Ninety percent of deaths from RMS occur within two years following diagnosis. For this reason, patients who survive free of disease beyond this point have a high probability of cure. Late relapses (beyond five years) occur in approximately 9 percent of patients [107].
Prognosis — Outcomes are variable for patients who relapse after initial therapy or whose tumors progress while on treatment. Recurrences are often, but not uniformly, fatal [108-110]. Five-year post-relapse survival rates in two large cooperative group series and a multicenter French series ranged from 17 to 49 percent [108,110-112]. Factors associated with a lower chance of cure after a first relapse include metastatic relapse, prior use of alkylating agents and RT, alveolar histology, shorter time to relapse, and higher stage and CG at diagnosis.
●In one report, based on 605 relapses observed in 2364 patients treated with IRS-IV pilot and IRS-IV protocols, prognosis was dependent upon initial histology, CG, and stage at diagnosis [108]. Patients with relapsed botryoid embryonal RMS had a 64 percent survival rate at five years, compared with 26 percent for patients with relapsed embryonal RMS and only 5 percent for those with relapsed alveolar RMS or undifferentiated sarcoma.
●In the second series, there were 1398 patients with localized disease at diagnosis who were treated within three consecutive European cooperative group protocols and who achieved a clinical remission [112]. In all, 474 had a relapse and 176 were alive ≥3 years from the last event. In multivariable analysis, the factors that were most strongly associated with poor outcomes were metastatic disease at relapse, prior RT treatment, initial tumor size >5 cm, and relapse within 18 months of diagnosis. Other significant factors included unfavorable site, nodal status, three- and six-drug treatment, and alveolar histology.
Treatment — As described above, outcomes with any regimen for relapsed patients are suboptimal, and the best regimen has yet to be identified. Local therapy (eg, surgery) should be considered if feasible [113]. (See "Surgical treatment and other localized therapy for metastatic soft tissue sarcoma".)
In an originally low-risk patient who was treated with VA only or short-course VAC, a new full-VAC (high-dose cyclophosphamide) regimen is appropriate and could be curative. Other options include combinations of doxorubicin, ifosfamide, and etoposide [114]; doxorubicin, ifosfamide, mesna, and dacarbazine [115]; cyclophosphamide plus topotecan; vinorelbine, cyclophosphamide, and temsirolimus; irinotecan plus vincristine, with or without temozolomide; alternating cyclophosphamide/doxorubicin and ifosfamide/etoposide; and others [111,116-124].
In patients with a very low chance for cure (relapsed high-risk or intermediate-risk patients treated with maximum standard therapy with RT for local control), we favor using regimens that will not affect quality of life significantly, including vinorelbine and cyclophosphamide (with or without temsirolimus), vinorelbine alone, or topotecan plus cyclophosphamide.
Long-term complications — A significant portion of children who are successfully treated for RMS develop long-term treatment-related sequelae. In a report of late complications from the IRS-II and III trials, short stature or facial asymmetry affected almost one-third, poor dentition affected one-fourth, and impaired vision, hearing, and learning affected one-sixth each [79].
Specific long-term follow-up guidelines after treatment of childhood cancer have been published by the COG and are available online [82].
Bowel and bladder problems — With modern therapy, bladder preservation rates as high as 60 percent are reported for patients with tumors of the bladder or prostate [7,32,125]. However, 25 percent of patients who retain their bladders suffer from significant bladder functional problems, including incontinence, frequency, and nocturia [126]. Moreover, patients treated for a bladder or prostate primary are at risk for bowel complications related to surgery or RT, hemorrhagic cystitis, and primary hypogonadism [126,127]. (See "Radiation proctitis: Clinical manifestations, diagnosis, and management" and "Chemotherapy and radiation-related hemorrhagic cystitis in cancer patients" and "Causes of primary hypogonadism in males".)
Dentofacial development — Sequelae of treatment for RMS involving the head or neck include several dentofacial abnormalities, including bony hypoplasia/facial asymmetry, trismus, tooth/root agenesis, microdontia, enamel defects, and osteonecrosis [128,129]. Meticulous long-term dental and radiographic follow-up are needed.
Cardiac dysfunction — Higher doses of cyclophosphamide may increase the risk for cardiac dysfunction [130]. (See "Cardiotoxicity of cancer chemotherapy agents other than anthracyclines, HER2-targeted agents, and fluoropyrimidines", section on 'Cyclophosphamide' and "Cancer survivorship: Cardiovascular and respiratory issues".)
Secondary tumors — Secondary tumors are a potentially devastating late complication of treatment [131-135]. Among 1770 patients treated in the IRS-I and II protocols, 22 secondary malignancies (11 postradiation bone sarcomas and 5 acute myeloid leukemias) developed at a median of seven years posttreatment [131]. Second cancers are more common in children treated with both chemotherapy and radiation [135]. (See "Radiation-associated sarcomas", section on 'Risk factors'.)
Endocrinopathies — In patients with head and neck RMS, RT to the hypothalamic-pituitary area can induce short stature secondary to growth hormone deficiency [136,137]. (See "Endocrinopathies in cancer survivors and others exposed to cytotoxic therapies during childhood".)
Eye problems — In one report, 82 percent of the 79 patients with orbital tumors who had received RT developed cataracts, while 59 percent had orbital hypoplasia [138].
TREATMENT OF ADULTS — We treat adults with RMS with the same protocols used in children. However, data suggest that compared with children, adults have inferior outcomes [77,78,139-144]. This difference is likely related to inadequacy of primary treatment in older patients and more advanced disease at presentation.
Given the rarity of RMS in adults, accounting for 2 to 5 percent of adult soft tissue sarcoma, most data about outcomes and treatment are retrospective [78,145]. As an example, in a compilation of data on outcomes of RMS among 1071 adults (age >19 years) and 1529 children derived from the Surveillance, Epidemiology, and End Results (SEER) database of the National Cancer Institute (NCI), adults had worse five-year survival (27 versus 61 percent) [139]. Tumors in adults were more likely to be at an unfavorable site (65 versus 55 percent) and to have histologies that are unusual during childhood, particularly the pleomorphic subtype and not otherwise specified. However, adults had overall worse survival than children for similar tumors. This was especially evident for localized disease (five-year survival 47 versus 82 percent).
Similarly, a retrospective analysis of over 2000 patients treated on Children's Oncology Group protocols for newly diagnosed RMS found that older patients (15 to 39 years old) were more likely to have factors associated with poor prognosis including primary tumors ≥5 cm, metastatic disease, alveolar histology, and a FOXO1 fusion [144]. In addition, both five-year event-free survival (44 versus 67 percent) and overall survival (52 versus 78 percent) were lower in older patients.
The poorer outcomes in adults may, in part, reflect the inadequacy of primary treatment, as evidenced by the following:
●In one report of 82 adults with locoregional RMS treated over a 38-year period at MD Anderson Cancer Center, the 10-year disease-free and overall survival rates were 41 and 40 percent, respectively [77]. However, only 71 percent received chemotherapy, and 11 percent were treated with radiation therapy (RT) alone.
●In a second series of 171 patients over age 19 who were diagnosed with RMS at a single institution over a 25-year period, five-year event-free survival (EFS) and overall survival rates were 28 and 40 percent, respectively [78]. However, among the 110 patients with embryonal, alveolar, or “not otherwise specified” RMS, the five-year survival rate was 61 percent for those who had high scores for appropriate treatment based upon modern treatment guidelines for pediatric RMS.
Given this data and the lack of any prospective studies about how to better modify therapy in older adults, we continue to treat our older patients according to the protocols used in children.
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Soft tissue sarcoma".)
SUMMARY AND RECOMMENDATIONS
●Overview of risk-adapted therapy
•Evolution of therapy – The treatment of rhabdomyosarcoma (RMS) has evolved considerably. Using contemporary combined modality therapy, over 70 percent of children with localized RMS can be cured. These improved outcomes are the direct result of the use of risk-based multimodality therapeutic protocols that have been developed by large international cooperative groups, such as the Intergroup Rhabdomyosarcoma Study Group (IRSG), also known as the Soft Tissue Sarcoma Committee of the Children's Oncology Group (COG).
•Treatment options – Modern treatment includes chemotherapy for primary tumor cytoreduction and eradication of both macroscopic and microscopic metastatic disease; surgery, if feasible; and radiation therapy (RT) to control microscopic local residual disease. The specific treatment regimen depends on the estimated risk of a disease recurrence, which is based upon a variety of clinicopathologic prognostic factors, an approach termed risk-adapted therapy (table 1 and table 4). This risk stratification system is used in North America. A different risk stratification system that does not incorporate molecular classification (fusion gene status) is used in Europe. (See 'Overview of risk-adapted therapy' above.)
●Treatment approach – Some general principles underlying treatment of RMS are outlined below. Treatment is constantly evolving, and if the reader's goal is to obtain information on how to treat a specific patient, he/she should contact a pediatric oncologist who is a member of the COG and who has expertise in soft tissue sarcoma for the latest recommendation. In general, adults with RMS are treated using the same principles as established for children. Results are most favorable when adults are treated in pediatric clinical trials, which are generally open to individuals up to age 50. (See 'Treatment of adults' above.)
•Surgery – We recommend complete excision for localized disease as long as functional and/or cosmetic results are acceptable (Grade 1A). If complete resection is not feasible or if disease involves the orbit, vagina, bladder, or biliary tract, patients are better approached with an initial diagnostic biopsy followed by induction (neoadjuvant) chemotherapy and then definitive local therapy. For disease at any primary site, palpable or radiologically enlarged (≥1 cm) lymph nodes should be excised and submitted for histologic evaluation, regardless of positron emission tomography (PET) avidity. The necessity of routine surgical sampling of nonsuspicious lymph nodes depends on the primary site. (See 'Role of surgery' above.)
•Radiation therapy – We recommend RT to enhance local control in all patients with RMS except those who have embryonal/fusion-negative clinical group (CG) I tumors (Grade 1B). Although the benefit of RT has yet to be determined, treatment standards in North America also recommend the addition of RT to all patients with CG I tumors that are fusion positive or are fusion negative with alveolar histology. (table 2). (See 'Clinical group I' above.)
•Chemotherapy – We recommend chemotherapy in addition to local therapy for all patients with RMS (Grade 1A). The choice of regimen depends on the estimated risk of a disease recurrence (table 4). (See 'Chemotherapy' above and 'Radiation therapy' above.)
●Clinical protocols – All eligible patients with RMS should be enrolled in one of the active clinical research protocols in any part of the world. In the United States, these are designed by the Soft Tissue Sarcoma Committee of the COG and administered by COG centers according to risk groups (table 1). For available open protocols, a member center should be contacted. Although in the United States pediatric oncologists usually treat patients up to the age of 20, any patient younger than 50 years old is eligible for these protocols. In particular, there is no standard approach for patients with high- and intermediate-risk disease at present. We suggest that all of these patients be enrolled in available national collaborative clinical trials or local clinical trials testing novel approaches. (See 'Treatment strategies according to risk group' above.)
For patients diagnosed in other parts of the world, cooperative group protocols are available through organizations such as the International Society of Paediatric Oncology (SIOP), the Weichteilsarkom Studiengruppe (CWS), and the European paediatric Soft tissue sarcoma Study Group (EpSSG).
●Long-term complications – A significant portion of children who are successfully treated for RMS develop long-term treatment-related sequelae. Specific long-term follow-up guidelines after treatment of childhood cancer have been published by the COG and are available online [82]. (See 'Long-term complications' above.)
●Recurrent disease – Recurrences are often, but not uniformly, fatal. Standard therapy for relapsed patients is suboptimal, and the best regimen has yet to be identified. There are several active chemotherapy regimens, but local therapy (eg, surgery) should be considered if feasible. (See 'Recurrent disease' above.)
ACKNOWLEDGMENTS — The UpToDate editorial staff acknowledges Marc Horowitz, MD, and Thomas F DeLaney, MD, who contributed to earlier versions of this topic review.
Do you want to add Medilib to your home screen?