Radiation Oncology/Rhabdomyosarcoma

Epidemiology

• 350 children per year in US
• Most common soft tissue sarcoma of childhood
• 2/3 occur <6 years, second peak mid-teen
• Clinical presentation depends on site, but frequently an asymptomatic mass
• Most common locations are H&N, GU, and extremities
• Local spread along fascial and muscle planes
• LN+ in 15%, incidence varies by subsite (ocular 1%, paratesticular 25%)
• DM+ in 15% at diagnosis, commonly lung, bone marrow, bone
• Li-Fraumeni syndrome: multiple tumors in siblings/relatives of children with RMS (adrenal, GBM, lung, breast, sarcomas)

Histology

• Myogenic precursor markers on IHC (actin, myosin, desmin, myoglobin, myoD, Z-band protein)
  • MyoD drives cells to myogenic lineage differentiation, and also stops cell replication. MyoD MAb available for diagnosis
• Two very distinct histologies and behavior patterns
• Embryonal (includes botyroid ERMS and spindle cell ERMS)
  • diploid to hyperploid DNA
  • no consistent transclocations identified, but loss of heterozygosity at 11p15.5
  • predominately in infants/young children
  • frequently H&N, GU location
  • considered to have favorable prognosis
• Alveolar
  • tetraploid DNA
  • translocations: t(2,13) PAX3-FKHR and t(1,13) PAX7-FKHR gene fusion, which regulates neuromuscular development
  • N-myc amplification (unlike embryonal)
  • predominately older children
  • predominately extremities, trunk, and retroperitoneum
• Undifferentiated
  • diffuse cell population of primitive, noncommitted mesenchymal cells
  • primarily a diagnosis of exclusion

Superior prognosis:

• Botyroid ERMS (6%) - 5-yr OS 95%
  • "grape like." Occurs in mucosa-lined organs -- bladder, vagina, nasopharynx, middle ear, biliary tree. Generally localized and non-invasive.
• Spindle cell ERMS (3%) - 5-yr OS 88%
  • Most common site is paratesticular.

Intermediate prognosis:

• Embryonal (59%) - 5-yr OS 66%
  • Orbit, GU, H&N
  • Blastemal mesenchymal cells. Resembles normal skeletal muscle of 7-10 wk old foetus. Often with LOH 11p15.5.

Poor prognosis:

• Alveolar (32%) - 5-yr OS 54%
  • Extremities, Trunk, Retroperitoneal
  • Line connective tissue septa reminiscent of alveoli. Resemble normal skeletal muscle of 10-20 wk old foetus. t(2;13) or t(1;13). Frequent n-myc amplification.
• Undifferentiated (1%) - 5-yr OS 40%

Prognostic factors

• Primary site - determines resectability (and thus IRS grouping)
• Tumor size (<5 or >5 cm)
• Lymph node involvement - +LN associated with more advanced primary
• Histology
• Complete resection

Staging and grouping

TNM Staging:

• T1 - Confined to anatomic site of origin
  • T1a - <= 5 cm
  • T1b - > 5 cm
• T2 - Extension beyond site of origin
  • T2a - <= 5 cm
  • T2b - > 5 cm

• N-stage: N0 or N1
• M-stage: M0 or M1

IRS Staging Classification (pre-treatment): (IRS-IV)

• Parameningeal H&N sites include: (IMMNNPPP) - Infratemporal fossa, Middle ear, Mastoid region, Nasal cavity, Nasopharynx, Paranasal sinus, Pterygopalatine fossa, +/- parapharyngeal region (sometimes included).

IRS Grouping (after resection):

Risk stratification

EMB = Embryonal, Botryoid, or Spindle cell
ALV = Alveolar or undifferentiated

• Low risk:
  • EMB Stage 1 (favorable site), Group I-III (non-metastatic)
  • EMB Stage 2-3 (unfavorable site), Group I-II (R0, R1, or LN+)

• Intermediate risk:
  • EMB Stage 2-3 (unfavorable site), Group III (gross disease)
  • ALV Stage 1-3 (any alveolar)

• High risk:
  • Metastases (Stage 4, Group IV)

Treatment

• Prior to 1970, RMS was treated with surgery alone or RT alone and had poor outcomes (OS <25%). Only exception was orbit and GU
• Series of trials: Intergroup Rhabdomyosarcoma Study Group IRS I-IV, V ongoing
• Risk adapted therapy based on: tumor size, site, histology, residual disease after surgery, LN involvement, distant mets, and patient age (see above)
  • 5-year OS improved from IRS-I 55% to IRS-II 63% to IRS-III/IV 71%
• Recommend treating all patients on protocol

IRS Trials

• IRSG-D9602 (1997-2004) - non-randomized.
  • Purpose: (1) decrease toxicity by reducing RT doses(2) eliminate cyclophosphamide in lowest-risk patients.
  • Subgroup A (lowest risk) - (embryonal, stage 1, group I/IIA; stage 1, group III orbit; or stage 2, group I) -- received VA (vincristine, dactinomycin)
  • Subgroup B - (embryonal, stage 1, group IIB/C; stage 1, group III non-orbit; stage 2, group II; stage 3, group I/II) -- received VA + cyclophosphamide
  • 342 pts (264 in subgroup A, 78 in subgroup B). No RT for Group I pts. Pts in group II-III received RT. In patients with stage I, group IIA disease, dose was reduced to 36 Gy (41.4 Gy in IRS-IV). For those with group III orbit disease, dose was reduced to 45 Gy (50-59 Gy in IRS-IV).
  • 2011 PMID 21357783 (Median f/u 5.1 yrs) -- "Results of the Intergroup Rhabdomyosarcoma Study Group D9602 Protocol, Using Vincristine and Dactinomycin With or Without Cyclophosphamide and Radiation Therapy, for Newly Diagnosed Patients With Low-Risk Embryonal Rhabdomyosarcoma: A Report From the Soft Tissue Sarcoma Committee of the Children's Oncology Group." (Raney RB, J Clin Oncol. 2011 Apr 1;29(10):1312-8)
    • 5-yr FFS 89% for subgroup A, 85% subgroup B. 5-yr FFS 81% (stage 1, group IIA), 86% (group III orbit)
    • Conclusion: Compared with historical controls (in IRS-III/IV, 5-yr FFS 83% and OS 95%), FFS and OS rates were similar to IRS-III including those with reduced RT doses, but inferior to IRS-IV pts with VA + cyclophosphamide. FFS rates were similar among subgroups A and B.

• IRS-IV (1991-1997)
  • Randomized. 883 patients with non-metastatic RMS after surgery. Treatment by tumor site, Group (1-3), and Stage (I-III)
  • CG III RT: conventional (50.4/28) vs. hyperfractionated (59.4/54 in 1.1 Gy BID). Also, WBRT omitted for all parameningeal unless CSF involvement documented
  • 2001 PMID 11408506 -- "Intergroup rhabdomyosarcoma study-IV: results for patients with nonmetastatic disease." (Crist WM, J Clin Oncol. 2001 Jun 15;19(12):3091-102.)
    • 3-year OS: 86%
    • CG III: HF-RT comparable to conventional RT
    • VAC and VAI or VIE equally effective
    • Embryonal RMS benefited from intense chemo, no difference for alveolar/undifferentiated
    • Prognostic subgroups developed (see above)

• IRS III (1984-1991)
  • Randomized. 1062 previously untreated patients. Treatment by clinical group (I-IV), histology (favorable-unfavorable), and site
  • RT: none in CG I favorable, CG I unfavorable and CG II 41.4 Gy, CG III 41.4 Gy if <5cm and <6 years, 50.4 Gy if >5cm or >6 years
  • 1995 PMID 7884423 -- "The Third Intergroup Rhabdomyosarcoma Study." (Crist W, J Clin Oncol. 1995 Mar;13(3):610-30.)
    • 5-year OS: 71% (compared with 62% in IRS II). 5-year PFS 65% (compared with 55% in IRS II)
    • CG I favorable: VA alone as good as VAC. 5-year PFS 83%
    • CG II favorable (excluding orbit, head, paratesticular): VA + RT comparable to VAD + RT
    • CG III (excluding orbit, head): intensive chemo better than VAC or VAC-VADRC. 5-year PFS 62%
    • CG III (bladder, vagina, pelvis): much better outcome due to early RT, with better bladder salvage (60% vs. 25%)
    • CG IV: no benefit from aggressive therapy
    • Whole brain RT omitted for parameningeal tumors with nerve palsy or base of skull erosion - risk of CNS relapse comparable

• IRS II (1978-1984)
  • Randomized. 999 previously untreated patients. Treatment by clinical group (I-IV), histology (favorable-unfavorable), and site
  • RT Dose escalation (CG I no RT, CG II 40-45 Gy, CG III and <5cm and <6 years 40-45 Gy, CG III and >5 cm or >6 years 50-55 Gy). WBRT if parameningeal tumors with nerve palsy, base of skull erosion, or intracranial extent
  • 1993 PMID 8448756 -- "The Intergroup Rhabdomyosarcoma Study-II." (Maurer HM, Cancer. 1993 Mar 1;71(5):1904-22.)
    • 5-year OS: 62% (compared with 55% in IRS I)
    • CG I (excluding extremity alveolar): VA comparable to VAC. Alveolar had high rate of relapse
    • CG II: intensive VA comparable to repetitive-pulse VAC
    • CG III (excluding pelvic): VAC comparable to VDC-VAC
    • Whole brain RT improved meningeal recurrence and increased survival in high risk patients with parameningeal sites (67% vs. 45% in IRS-I)
    • CG IV: no benefit over IRS-I

• IRS I (1972-1978)
  • Randomized. 686 previously untreated patients. Minimum F/U 7 years
  • 1988 PMID 3275486 -- "The Intergroup Rhabdomyosarcoma Study-I. A final report." (Maurer HM, Cancer. 1988 Jan 15;61(2):209-20.)
    • 5-year OS: 55%, after relapse 32% at 1 year. Risk of DM much greater than LR
    • CG I: VAC, no benefit to RT. Subsequent analysis showed benefit in alveolar or undifferentiated histology.
    • CG II: No difference between VA + RT or VAC + RT
    • CG III and IV: No difference between VAC + RT or VACD + RT. 5-year OS CG III 52% vs. CG IV 20%
    • No dose-response relationship 40-60 Gy
    • No difference in RT to whole muscle bundle vs. tumor + margin only
    • Parameningeal sites with high risk factors (palsy, erosion of base of skull, etc) had high incidence of CNS relapse

Surgery

• Only ~20% of patients can undergo R0 resection, and another 20% can undergo compromised R1 resection
• Trend toward less aggressive surgical resection and more reliance on chemo-RT
• Normal margin of 5mm typically required (Perez 5th edition) to be considered complete excision

Radiotherapy

General:

• RMS infiiltrates along tissues planes, and may extend beyond obvious visible margins
• Treat pre-chemo, pre-surgery tumor volume. 2 cm margin is sufficient (per IRS-IV)
• For parameningeal sites, treatment of adjacent meninges can prevent meningeal relapse
• Elective LN irradiation not necessary if clinically negative and will be treated with chemotherapy (Perez 5th ed)
• Hyperfractionated BID RT does not improve outcomes in a randomized setting
• Timing (newly-revised): Low risk - week 13 of chemo. Intermediate risk - week 4 of chemo. High risk - week 20 of chemo. Day 0 for documented intracranial extension.

IRS-V (ongoing):

• low risk disease
  • complete resection, SM- = no RT
  • microscopic SM+ and LN- = 36 Gy
  • microscopic SM+ and LN+ = 41.4 Gy
  • orbital primary with residual gross disease = 45 Gy
  • non-orbital primaries with residual gross disease = 50.4 Gy
• intermediate risk disease
  • complete resection, SM- = 36 Gy
  • miscroscopic SM+ = 41.4 Gy
  • gross residual disease = 50.4 Gy

• RT guidelines
  • CTV = pre-chemo GTV + 1.5 cm
  • PTV = CTV + institutional margin (0.3 - 0.5 cm)
  • Since this margin (1.8 - 2cm) is larger than the original IRS margin (2cm from block, which translates to effective 1.5cm due to penumbra), next generation of protocols will have CTV - PTV expansion of 1.5 cm
  • Start time: Day 0 if intracranial extension, week 13 for low risk, week 4 intermediate risk, week 20 high risk (primary site) and week 47 (metastases)

Chemotherapy

• Chemotherapy is used for all patients
• Non-metastatic: VAC (vincristine, actinomycin D, cyclophosphamide)
• Metastatic: Currently using "up-front window" approach:
  • Test 1-2 agents for 1-2 cycles during 6 weeks
  • If they respond, add to standard VAC + XRT

• Topotecan
  • Metastatic disease: highly active
• Melphalan
  • Metastatic disease: inferior survival compared to IE
• Ifosfamide/Etoposide
  • Recurrent disease: active
  • Local disease: comparable outcome VAC vs. VAI vs. VIE, but toxicity and logistics better for VAC
• Cisplatin/Etoposide
  • No benefit VAC/dox/cisplatin and VAC/dox/cisplatin/etoposide vs. VAC alone
• Doxorubicin
  • No benefit VAC/dox vs. VAC alone
• Base chemotherapy is VAC (vincristine, actinomycin D, cyclophosphamide)

Specific Subsets

Clinical Group I (complete resection)

• IRS-I/II/III, 1999 (1972-1991) PMID 10550144 -- "Indications for radiotherapy and chemotherapy after complete resection in rhabdomyosarcoma: A report from the Intergroup Rhabdomyosarcoma Studies I to III." (Wolden SL, J Clin Oncol. 1999 Nov;17(11):3468-75.)
  • Retrospective analysis of IRS-I to IRS-III protocols. 439 patients with completely resected RMS (Group I), treated with VAC +/- dox/cisplatin. 19% received RT
  • 10-year outcomes: OS 89%, FFS 79%. Failures: 6% local, 6% regional, 7% distal
  • ERMS: no difference in OS, trend to benefit for FFS
  • ARMS/undifferentiated:
    • IRS I/II: 10-year OS RT+ 82% vs. RT- 52% (SS), 10-year FFS 73% vs. 44% (SS)
    • IRS III: 10-year OS RT+ 95% vs. RT- 86% (NS), 10-year FFS 95% vs. 69% (SS)
  • Conclusion: Patients with Group I embryonal RMS have excellent prognosis without RT; alveolar RMS outcomes substantially improved with RT

Head & Neck

• 35% of RMS patients (15% parameningeal, 10% orbit, 10% other)
• Majority are unresectable (CG III) at diagnosis
• Orbit and non-parameningeal sites have high cure rates
• Parameningeal sites (see above) have worse control
  • Bad prognostic factors: base of skull erosion, intracranial extension, cranial nerve palsy, primary location in paranasal sinuses or pterygoid/infratemporal fossa
  • Local failure in IRS-IV was 16%, with DM 11%
  • Local failure from MSKCC series was 5%, and 0% in CG I-III as reported in IRS-IV

• SIOP MMT 89 & 95 1989-2003) PMID 19204197 -- "Treatment of nonmetastatic cranial parameningeal rhabdomyosarcoma in children younger than 3 years old: results from international society of paediatric oncology studies MMT 89 and 95." (Defachelles AS, J Clin Oncol. 2009 Mar 10;27(8):1310-5. Epub 2009 Feb 9.)
  • Prospective. 59 children, cranial parameningeal RMS. Goal to avoid RT if CR after chemo
  • Outcome: 5-year EFS if RT used 59% vs. if no RT used 28% (SS). Only 12% ultimately cured without RT
  • Conclusion: Despite concerns about late effects, cure of PM RMS remains unlikely without RT

• MSKCC, 2005 PMID 15817347 -- "Intensity-modulated radiotherapy for head-and-neck rhabdomyosarcoma." (Wolden SL, Int J Radiat Oncol Biol Phys. 2005 Apr 1;61(5):1432-8.)
  • Retrospective. 28 patients (21 parameningeal, 3 orbital, 4 other), treated with IMRT. 89% Group III disease, 71% intracranial extension . Median tumor size 5.2 cm. RT median dose 50.4 Gy (30-55.8), 1.5 cm margin (1cm CTV + 0.5 PTV). No boost.
  • 3-year OS: 65%, 3-year FFF 95% locally (100% for Groups I-III), 90% regional LN, 88% CNS, 80% distal sites. Alveolar RMS significantly worse
  • Toxicity: acute comparable to IRS trials, late was mild (1 cataract, 1 GH deficiency, 1 mild facial asymetry)
  • Conclusion: Outstanding local control despite reduced margin

• IRS II-IV, 2004 (1978-1997) PMID 15234036 -- "Influence of radiation therapy parameters on outcome in children treated with radiation therapy for localized parameningeal rhabdomyosarcoma in Intergroup Rhabdomyosarcoma Study Group trials II through IV." (Michalski JM, Int J Radiat Oncol Biol Phys. 2004 Jul 15;59(4):1027-38.)
  • Retrospective. 595 patients with PM-RMS registered. 95% with Group III disease. RT varied per protocol (40, 50.4, 59.4 BID). 21% patients (40% on IRS-II) received WBRT
  • 5-year OS: 73%. 5-year LF: 17%
  • Intracranial extension: increased 24% to 41% with cross-sectional imaging
  • Timing of RT:
    • Meningeal impingement: <2 weeks after chemo 18% LF vs. >2 weeks 33% (SS)
    • Intracranial extension: <2 weeks after chemo 16% LF vs. >2 weeks 37% (NS, p=0.07)
    • Cranial nerve palsy/ base of skull bone erosion: <2 weeks after chemo 21% LF vs. >2 weeks 30% (NS)
    • If no signs of meningeal impingement: no difference in LF if RT start earlier/later than 10 weeks
  • RT Dose
    • Large (>=5 cm) tumors: <47.5 Gy LF 35% vs. >47.5 Gy 15%. Small tumors no difference
    • WBRT no impact on LR or CNS failure