Rhabdomyosarcoma

RMS is broadly classified as one of the small, round, blue cell malignancies of childhood.⁷ The histologic subtypes, in order by worsening prognosis, include embryonal rhabdomyosarcoma (ERMS), the most common form (with spindle cell and botryoid comprising two favorable variants), alveolar rhabdomyosarcoma (ARMS), and undifferentiated sarcoma. Patients with undifferentiated sarcoma were previously enrolled on the IRSG clinical trials in a fashion similar to ARMS. Currently, these patients receive systemic therapy similar to patients with Ewing’s sarcoma. Light microscopy often describes ERMS with spindle-shaped cells and ARMS with small, round, blue cells forming alveolar-like spaces, although tumor biopsy specimens may also look like collections of poorly differentiated cells complicating definitive diagnosis by hematoxylin and eosin staining alone. Immunohistochemical staining can include positivity for MyoD.⁸ Cytogenetic events occur in ERMS with a loss of heterozygosity on chromosome 11p and in ARMS with translocations between chromosomes 2 and 13 or 1 and 13. These represent loci of PAX3 and PAX7 (chromosomes 2 and 1) and their fusion with FKHR on chromosome 13. These fusion proteins act as transcription factors and are diagnostic of ARMS.⁷ Other genetic abnormalities and syndromes associated with RMS include neurofibromatosis type 1 (with a prevalence of 1 : 200 noted on IRSG-IV), Li-Fraumeni syndrome, and Beckwith-Wiedemann syndrome.^9–11

Nonrhabdomyosarcoma Soft Tissue Sarcoma

The pediatric NR-STS comprise a heterogeneous group of tumors managed homogeneously owing primarily to the limitations in both patient numbers and effective chemotherapy. The most common histologies noted in practice and clinical trials include synovial cell sarcoma and malignant peripheral nerve sheath tumor (MPNST), although several rarer variants including rhabdoid tumor and infantile fibrosarcoma are almost exclusive to the pediatric age group.^12,13 Although many NR-STSs have no genetic signature, several histologic variants do have specific translocations or deletions^12,14 (Table 66-1).

TABLE 66-1 Genetic Abnormalities: Nonrhabdomyosarcoma Soft Tissue Sarcomas

Rhabdomyosarcoma

The presentation of children with RMS is diverse and relates to the site of involvement and extent of disease. The median age for children presenting with RMS is younger than 5, although another peak occurs in the mid teens. The head and neck region is the most frequent site of involvement and is divided into favorable and unfavorable (parameningeal) sites.^15,16 This site comprised 42% of localized presentations in patients enrolled on IRSG III and IRSG IV and serves as an example of the important nature of primary tumor location.¹⁷ Metastatic disease occurs 20% of the time at presentation, and nodal involvement is rare (<5%) except for paratesticular (25%) and extremity (24%) sites of disease, which warrant either computed tomography (CT) (paratesticular, <10 years of age), nodal sampling (paratesticular, ≥10 years of age), or sentinel node biopsy (extremity site).^18–21

Nonrhabdomyosarcoma Soft Tissue Sarcoma

The presentation of NR-STS is often as a painless mass whereas other symptoms are site specific. Approximately half of the cases arise in the extremity, and the incidence increases throughout the adolescent years and into adulthood.^22,23 Although the vast majority of soft tissue “masses” are either not real or benign, consideration should be given to malignancy because an ill-chosen surgical approach may compromise future local therapy (Fig. 66-1).

Figure 66-1 Rhabdomyosarcoma of the pinna of the ear presenting as a painless soft tissue mass.

Fifteen to 22 percent of patients present with metastatic disease at diagnosis, with the lungs being the predominant site of metastatic involvement.^13,24 Clear cell sarcoma, epithelioid sarcoma, and angiosarcoma carry a risk of regional nodal involvement and warrant evaluation of the draining nodal bed(s) with sentinel node sampling.

Rhabdomyosarcoma

Staging workup is similar for most patients with RMS (Table 66-2). A site-specific history and physical examination should be performed by the treating radiation oncologist at the time of presentation to define the site of primary disease, its extent, and symptoms related to potential metastatic sites of involvement. Attention to cranial nerve involvement for head and neck primary sites as well as attending the examination under anesthesia for genitourinary sites (vaginal, cervix, uterus, bladder, and prostate) will assist in the multidisciplinary discussion regarding staging and local therapy approaches.

TABLE 66-2 Staging Workup and Follow-Up Evaluations for Patients with Rhabdomyosarcoma and Nonrhabdomyosarcoma Soft Tissue Sarcomas

Laboratory studies include a complete blood cell count and chemistry panel, which may be suggestive of bone marrow disease (anemia) or other organ-specific dysfunction related to metastatic involvement.

Review of the pretreatment imaging is critical to ensure both adequate imaging of the entire primary tumor as well as understanding the extent of disease. Often, standard diagnostic imaging protocols (particularly magnetic resonance imaging [MRI]) will not encompass the entire tumor, because the studies were designed for evaluation of joint or intracranial disease. Repeating these studies will facilitate better local therapy decisions and delivery. Standard diagnostic imaging should include MRI of the primary site of disease, CT of the chest (and abdomen for infradiaphragmatic primary tumors), bone scintigraphy, and site-specific imaging related to concern for metastatic disease. ¹⁸F-Fluorodeoxyglucose-labeled positron emission tomography (FDG-PET) may be helpful and could potentially replace bone scintigraphy, but current reimbursement may limit its routine application.

Bone marrow aspiration and biopsy should be performed for all patients, as well as cerebrospinal fluid cytology for patients with parameningeal, orbital, and paraspinal primary sites of disease. Lymph node sampling should be performed for patients 10 years of age or older with paratesticular primary tumors, and sentinel node mapping and sampling should be done for patients with disease of an extremity.

After staging evaluation, a clinical stage should be assigned (outlined in Table 66-3). If surgery has been performed (or after a planned surgical resection or biopsy), a grouping should be defined (Table 66-4). These two factors plus the patient’s histology are combined into a risk classification to assign patients to protocol therapy.

TABLE 66-4 Clinical Grouping System for Rhabdomyosarcoma

Nonrhabdomyosarcoma Soft Tissue Sarcoma

The workup for patients with NR-STS is similar to that for RMS (see Table 66-2). A site-specific history and physical examination should be performed by the treating radiation oncologist at the time of presentation to define the site of primary disease, its extent, as well as symptoms related to potential metastatic sites of involvement.

Laboratory studies include a complete blood cell count and chemistry panel, the findings from which may be suggestive of organ-specific dysfunction related to metastatic involvement.

Diagnostic imaging studies include MRI of the primary site, CT of the chest, and other metastatic imaging, including bone scintigraphy, as indicated by symptoms and laboratory studies. FDG-PET may be useful in the workup of patients but is not currently considered standard of care.

Rhabdomyosarcoma

Favorable and unfavorable sites of disease are defined as part of the staging system and are noted in Table 66-3. In addition, histology affects outcome and is related to site of disease; favorable sites of disease in the orbit and genitourinary region are four and five times as likely to be of an embryonal histology as opposed to alveolar or undifferentiated histology. This yields higher disease control rates for localized ERMS compared with ARMS (failure-free survival [FFS], 82% vs. 65%).¹⁷ Clinical group (see Table 66-4) is also prognostic of FFS in both ERMS and ARMS. Because of the strong interaction of site, histology, group, and stage, a risk classification system was devised to stratify patients for therapeutic decisions. This resulted in three strata^17,25,26:

Nonrhabdomyosarcoma Soft Tissue Sarcoma

Stratification for patients with NR-STS is currently based on retrospective data being tested in a prospective clinical trial.^22,23 Low-risk patients are considered as having low-grade tumors (POG grade I or II) as well as high-grade tumors (POG grade III) less than or equal to 5 cm with a 5-year overall survival (OS) of 89%. Intermediate-risk patients with large (>5 cm) high-grade tumors have a 5-year OS of 56%, and high-risk patients with metastatic disease fare poorly, with a 5-year OS of 15%.²²

Rhabdomyosarcoma

All patients with RMS are managed in a combined modality fashion. Participation by a pediatric oncologist, pediatric surgeon, radiation oncologist, and orthopedic oncologist (if indicated) at the outset ensures optimal planning for the many therapeutic decisions necessary for this diverse and complicated disease. This multimodal approach has been the hallmark of the studies of the IRSG performed in North America. Since 1972, these trials have enrolled thousands of patients, unified the therapeutic approach across the United States, and improved OS for this disease.

IRSG I (1972–1978) and IRSG II (1978–1984) answered several important questions. Group I (without RT) and group II patients (with RT at doses of 40 to 45 Gy) were noted to fare well with a VA regimen (vincristine, dactinomycin [Actinomycin], omitting cyclophosphamide).^15,27 Group III patients did not benefit from the addition of doxorubicin, whereas patients with parameningeal disease treated without directed meningeal RT (a 2-cm meningeal margin) fared significantly worse (OS, 45% vs. 67%).¹⁵ Unfortunately, 75% of group IV patients continued to experience progressive disease.

IRSG III (1984–1991) went on to confirm that the VA regimen is equivalent to therapy with VAC and resulted in a progression-free survival of 83% without RT. The addition of doxorubicin to VA did not improve outcome for group II disease, and patients with group III/IV disease did not benefit from the addition of cisplatin or etoposide, with progression-free survival for the latter group remaining at 27%.²⁸

IRSG IV (1991–1997) was the first trial to employ presurgical staging in addition to the grouping system. A randomized question was asked about systemic therapy (now standard VAC vs. VAI [ifosfamide] vs. VIE [ifosfamide/etoposide]) for patients with stage I to III disease. Despite the hope for improvement in overall outcomes, no benefit was seen, and VAC remained the standard of care.²⁹ Patients with group III disease also underwent a randomization between standard RT (50.4 Gy/1.8 Gy daily fraction) and hyperfractionated RT (59.4 Gy/1.1 Gy twice daily fractions). Hyperfractionated RT yielded the same FFS (73%) as conventional RT, and local failure was equivalent (15% vs. 12%).³⁰ IRSG IV, in a retrospective fashion compared with IRSG III, also clarified the management of patients with group I paratesticular disease. Patients younger than age 10 years could be evaluated by CT for retroperitoneal lymphatic involvement (without retroperitoneal lymph node dissection [RPLND]), maintaining a 90% FFS. Those 10 years old or older benefited from RPLND (100% FFS with RPLND compared with 63% with CT alone) to define nodal involvement and the need for subsequent lymphatic RT.²⁹

IRSG V is the last of the intergroup named trials, now named with the COG nomenclature ARST. Patients with group II disease had a reduction in RT dose from 41.4 Gy to 36 Gy for positive surgical margins. Orbital primary tumors were also irradiated using a reduced dose of 45 Gy. Outcome on this low-risk trial appears equivalent to earlier studies, suggesting this is a safe and effective approach.³¹ Another randomized question was tested for intermediate-risk patients comparing VAC against VAC alternated with VTC (substituting topotecan). This did not improve outcome, with a 4-year FFS of 73% and 68%, respectively.³² High-risk patients during this era underwent phase II window therapy trials incorporating irinotecan. Activity of this agent in combination with vincristine was documented, but FFS remained poor (26% at 2 years).²⁵

The current round of trials in low risk RMS (ARST0331) further tailors and reduces therapy for patients with ERMS, resected upfront (group I/II), stage 1 to 3 and unresected (group III), stage 1 disease. The RT dose remains 36 Gy for microscopic disease and 41.4 Gy for nodal disease. Intermediate-risk patients (ERMS stages 2/3, group III and alveolar RMS stage 1 to 3, group I to III) are randomized between VAC and VAC/VI (ARST0531), incorporating irinotecan from the prior high-risk pilot trials. High-risk patients receive multiple-agent systemic therapy (ARST0431) and are also eligible for trials incorporating an insulin-like growth factor-1 receptor inhibitor (ARST08P1), seeking improvement in FFS in this poor prognosis population.

A general algorithm for the integration of chemotherapy, surgery, and RT is shown in Figure 66-2. As a first principle in nearly all North American RMS trials, only a nonmorbid surgical resection should be performed. Amputation, exenteration, or other extensive procedures are discouraged because RT will offer equivalent rates of local control with organ preservation. The timing of RT has varied across studies and disease stage. RT may be integrated after multiple weeks of chemotherapy, facilitating the time necessary to plan with modern radiotherapeutic techniques, allowing some reduction in tumor bulk, as well as potentially integrating a second-look surgical procedure. One exception to this variance is the consistent early introduction of RT in patients with parameningeal disease and intracranial extension.³³

Figure 66-2 General algorithm for the delivery of multimodal therapy for rhabdomyosarcoma.

Nonrhabdomyosarcoma Soft Tissue Sarcoma

The management of children with NRSTS, particularly the local therapy component, has paralleled that of adult paradigms, including the move to wide local surgical excision (WLE) from amputation, the role of adjuvant RT, and, most recently, the shift to preoperative chemoirradiation. Compared with RMS, clinical trials in North America for children with NR-STS have been smaller, infrequent, and have not resulted in well-defined prognostic groups from either a local therapy or overall outcome perspective.

The mainstay for management of children with NR-STS is surgery. Absent this modality, the outcome for patients is poor, with reported event-free survival (EFS) of 33% to 40% at 5 years.^23,34 The presence of metastatic disease also predicts for subsequent disease recurrence in the majority of patients (10% EFS).³⁴ All patients, but particularly those with resectable primary site disease, benefit from multidisciplinary discussion before initiation of therapy. Planning the appropriate surgery in the context of tumor grade, size, location, probable surgical margin status, and the patient’s age is the first critical step in the local management of patients with NR-STS. No prospective, completed clinical trials exist in the pediatric population to assist clinicians in selecting patients for specific aspects of local or systemic chemotherapy. The ongoing COG trial (ARST0332), stratified by risk group and local therapy approach, should accrue approximately 400 patients and determine the relative benefit of specific local and systemic therapeutic approaches. This multimodal approach to therapy incorporates standard-of-care approaches for NR-STS adapted for children and is outlined in Figure 66-3. More importantly, this trial will provide a uniform, prospectively treated population of patients on which to evaluate prognostic factors, both clinical and biologic. This could lead to improved risk stratification of patients with NR-STS on future trials.

Figure 66-3 General algorithm for the delivery of multimodal therapy for nonrhabdomyosarcoma soft tissue sarcomas.

Indications and Outcomes of Local Therapy

Rhabdomyosarcoma

The role of RT is to eradicate microscopic disease present after resection, treat gross disease, and consolidate visible sites of metastatic involvement. The indications and doses, although complicated by site, can be generalized (Table 66-5). Doses are delivered at 1.8 Gy per fraction daily.

TABLE 66-5 Indication for Radiation Therapy and Prescribed Doses for Patients with Rhabdomyosarcoma

Disease Status	Embryonal Histology	Alveolar Histology
Margin negative	No radiation	36.0 Gy
Margin positive	36.0 Gy	36.0 Gy
Node positive	41.4 Gy	41.4 Gy
Gross disease	50.4 Gy	50.4 Gy

Data for the treatment of completely resected disease (group I) come from review of prior IRSG trials I to III demonstrating improvement in FFS and OS for patients with ARMS when RT is delivered (73% vs. 44% and 82% vs. 52%, respectively).³⁵ Patients with group II disease received adjuvant radiation, resulting in local failure rates of less than 10%.³⁶ Patients with group III disease have local failure rates of 10% to 15%, depending on site, size, and era of treatment.^30,37,38

The parameningeal sites (middle ear, nasopharynx, paranasal sinuses, infratemporal and pterygopalatine fossae, and parapharyngeal region) require special attention to the skull base for adequate coverage of possible intracranial extension. If the timing of RT is not within the first 4 weeks (by protocol design) and intracranial extension is present (Fig. 66-4), then RT should be initiated within the first few weeks of the start of chemotherapy.

Figure 66-4 Coronal T1 contrast-enhanced MR image obtained at presentation in a child with intracranial extension from a pterygopalatine fossa rhabdomyosarcoma.

The orbit is a favorable site of disease involvement. Local management for this site with RT calls for delivery of 45 Gy at 1.8 Gy daily and yields excellent local control.³¹ Despite the extremely thin and fragile nature of the bony orbit and the adjacent globe, these two structures are essentially never breached by orbital primary tumors and should be constrained out of the clinical target volume. If tumor does extend through these structures, consideration should be given that this may be parameningeal instead.

Female genitourinary tract (vulvar, vaginal, cervix, and uterus) tumors of ERMS (particularly the botryoid subtype) require adequate local therapy despite the often young age of the patients (frequently <2 years of age [Fig. 66-5]). These patients should either undergo complete resection (group I) at diagnosis or receive adjuvant RT with either external beam or interstitial/intracavitary techniques. Absent adequate local control, high failure rates have been noted in the most recent low-risk clinical trials, prompting the call for improved local control.

Figure 66-5 Botryoid embryonal rhabdomyosarcoma of the female urogenital tract in an infant.

Paratesticular RMS is initially managed with radical inguinal orchiectomy with high ligation of the spermatic cord. RPLND, following a surgical template based on laterality, is conducted for all children 10 years of age or older.¹⁹ Younger patients may still be managed with CT evaluation of their retroperitoneal lymphatics. Treatment of the lymphatic region is based on the discovery of microscopically or grossly involved lymph nodes.

Techniques

All treatment for pediatric patients with soft tissue sarcomas should be volumetric imaging based, require target volume and normal tissue delineation, and allow plan evaluation with dose-volume histograms. Conformal, intensity-modulated, and proton beam RT are all acceptable methods of radiation delivery backed by published outcome data from both cooperative group and institutional clinical trials. Initial and postchemotherapy imaging studies are used to define the extent of disease. MRI is essential because CT provides poor soft tissue definition for evaluating tumor extent. The definitions for gross tumor volume and clinical target volume are described in Table 66-6 for both RMS and NR-STS. Target volume delineation is anatomically constrained, meaning that the radiation oncologist determines the infiltrativeness of the gross tumor toward adjacent normal tissues when designing the shape of the clinical target volume. Planning volumes are institution, site, and treatment specific (e.g., photons compared to protons), but are typically 3 to 5 mm with some form of daily localization (cone-beam CT or fiducials) using photon techniques. Motion may be defined and managed through respiratory gating as necessary.

TABLE 66-6 Target Volume Delineation for Rhabdomyosarcoma and Nonrhabdomyosarcoma Soft Tissue Sarcomas

Treatment planners should consider adjacent normal tissues and weigh the relative risks of site-specific late effects relative to one another. Every effort should be made to deliver the recommended prescribed dose of radiation because doses utilized in many pediatric malignancies have already been lessened significantly because of concerns of late effects. Parallel-opposed fields should be avoided in nearly all cases to reduce normal tissue exposure to high doses of radiation. Even in extremity sites, treatment with circumferential radiation using an intensity-modulated RT technique (resulting in low doses to the entire circumference of the extremity as opposed to sparing a strip of skin [Fig. 66-6]) has not resulted in clinically documented lymphedema with moderate follow-up.¹³

Figure 66-6 Cross-sectional image of an intensity-modulated RT plan for alveolar rhabdomyosarcoma of an extremity.

Interstitial brachytherapy may be used for NR-STS and RMS (less frequently) at the time of surgical resection. Catheters are usually placed into the operative bed transcutaneously while the resection cavity is exposed after removal of the primary tumor. Catheters are spaced at 1-cm intervals and extend 1 to 2 cm beyond the known tumor bed. The tumor bed may be marked with surgical clips or other radiopaque markers to define the region to be treated at the time of planning. High-dose-rate brachytherapy is delivered to a dose of 34 Gy at 3.4 Gy twice daily, and low-dose-rate brachytherapy is delivered to a dose of 40 to 45 Gy at approximately 9 Gy daily.

Timing