Clinical group and modified TNM stage for rhabdomyosarcoma: a review from the Children’s Oncology Group

Abstract

The Children’s Oncology Group (COG) uses Clinical Group (CG) and modified Tumor Node Metastasis (TNM) stage to classify rhabdomyosarcoma (RMS). CG is based on surgicopathologic findings and is determined after the completion of initial surgical procedure(s) but prior to chemotherapy and/or radiation therapy. The modified TNM stage is based on clinical and radiographic findings and is assigned prior to any treatment. These systems have evolved over several decades. We review the history, evolution, and rationale behind the current CG and modified TNM classification systems used by COG for RMS. Data from the seven most recently completed and reported frontline COG trials (D9602, D9802, D9803, ARST0331, ARST0431, ARST0531, ARST08P1) were analyzed and confirm that CG and modified TNM stage remain relevant and useful for predicting prognosis in RMS. We propose updates based on recent data and discuss factors warranting future study to further optimize these classification systems.

Introduction

Rhabdomyosarcoma (RMS) is the most common soft tissue sarcoma of childhood.¹ The Children’s Oncology Group (COG) utilizes two classification systems for patients with newly diagnosed RMS: Clinical Group (CG) and modified Tumor Node Metastasis (TNM) Stage.^2-4 Importantly, the modified TNM system used by COG for RMS is distinct from the American Joint Committee on Cancer (AJCC) TNM system maintained by the AJCC and Union for International Cancer Control (UICC). Thus, for clarity, this review will refer to the modified TNM system used by COG for RMS as the “COG RMS modified TNM staging system.” Herein, we describe the rationale for using CG and COG RMS modified TNM stage, their evolution, and their current definitions. Additionally, we analyze the prognostic value of these systems and propose changes for consideration.

CG and COG RMS modified TNM stage have played a key role in risk stratification for RMS. In 1972, three cooperative cancer study groups - the Cancer and Leukemia Group B, the Children's Cancer Study Group, and the Southwest Oncology Group - formed the Intergroup Rhabdomyosarcoma Study Group (IRSG).³ The IRSG conducted serial prospective therapeutic studies in RMS. After the formation of the COG in 2000, the IRSG efforts were continued through the COG Soft Tissue Sarcoma (COG-STS) Committee.⁵ These studies include: IRS I (1972-1978); IRS II (1978-1984); IRS III (1984-1991); IRS IV pilot studies (1987-1991); IRS IV series (1991-1997), IRS V also known as the D-series of studies (1996-2005); and the ARST series (2002-current).^3,6-21 We discuss CG and COG RMS modified TNM stage in the context of these trials.

The staging work-up for RMS has evolved over time (see Supplemental Document 1) and includes clinical examination, radiologic assessment of the primary site and regional lymph nodes, evaluation of surgical margins following upfront resection(s), and for certain clinical indications may include a chest CT, a radionucleotide bone scan or whole body 18-fluorodeoxyglucose (18-FDG)-Positron Emission Tomography (PET)-scan, bilateral bone marrow aspirates and biopsies, biopsy of regional lymph nodes and suspected metastatic sites, and lumbar puncture.²² The appropriate work-up is guided by each individual patient’s disease characteristics.²² This information is used to determine CG and COG RMS modified TNM stage.

Methods

Data from 2157 evaluable patients with newly diagnosed RMS in the COG database were analyzed. These patients were enrolled on seven previously reported frontline COG trials between 1997 to 2013 (D9602, D9802, D9803, ARST0331, ARST0431, ARST0531, ARST08P1). Event free survival (EFS) and overall survival (OS) were calculated using the Kaplan-Meier method.²³ Chi-squared test was used to assess the association between tumor size and tumor invasiveness. To find the optimal size cut point, data was analyzed using the named Contal and O’Quigley method.²⁴

Clinical Group

Overview and history of the Clinical Group system

The CG classification system was developed by the IRSG in 1972 to facilitate treatment assignment in the first IRS study, IRS I.^3,4,25 At that time, extent of disease at diagnosis and resectability were recognized as prognostic features in RMS.^26,27 Thus, the CG system classifies a patient’s disease status as one of four clinical groups, signified by Roman numerals I-IV, based upon the degree of disease spread at diagnosis and amount of disease remaining after the initial surgical procedure(s).⁴ CG is determined prior to initiation of chemotherapy and radiation, and does not change thereafter, regardless if delayed surgical resection occurs. The current system is shown in Table 1. While CG is considered a surgical-pathological system, in practice, clinical examination and imaging findings often contribute, particularly to the assignment of Group IV.

TABLE 1:

Current and Updated Clinical Group Classification System

Clinical Group	Current Classification	Our Updated Classification
Clinical Group I	Localized disease, completely resected a. Confined to muscle or organ of origin b. Contiguous involvement – infiltration outside the muscle or organ of origin, as through fascial planes. Note: Gross inspection and microscopic confirmation of complete resection is required. Note: Regional nodes not involved – lymph node biopsy or sampling is highly advised, except for head and neck lesions. Any nodes that may be inadvertently taken with the specimen must be negative.	Localized disease, completely resected Note: Gross inspection and microscopic confirmation of complete resection is required. Note: Regional nodes not involved.
Clinical Group II	Total gross resection with evidence of regional spread a. Grossly resected tumor with microscopic residual disease. Surgeon believes that he has removed all of the tumor, but the pathologist finds tumor at the margin of resection and additional resection to achieve clean margin is not reasonable. No evidence of gross residual tumor. No evidence of regional node involvement. b. Regional disease with involved nodes, completely resected with no microscopic residual. In contrast to Clinical Group IIa, regional nodes are involved, but completely resected and the most distal node is histologically negative. c. Regional disease with involved nodes, grossly resected, but with evidence of microscopic residual and/or histologic involvement of the most distal regional node in the dissection.	Localized disease, grossly resected with microscopic residual disease or regional disease, grossly resected with or without microscopic residual disease  a. Localized disease, grossly resected tumor with microscopic residual disease, regional nodes not involved  b. Regional disease with involved nodes, completely resected with no microscopic residual (including most distal node is histologically negative)  c. Regional disease with involved nodes, grossly resected with evidence of microscopic residual and/or histologic involvement of the most distal regional node in the dissection
Clinical Group III	Incomplete resection with gross residual disease a. After biopsy only b. After gross major resection of the primary (>50%)	Localized or regional disease, biopsy only or incomplete resection with gross residual disease
Clinical Group IV	Distant Metastatic disease present at onset (Lung, liver, bones, bone marrow, brain, and distant muscle and nodes) The following are also considered evidence of metastatic disease and place the patient in Group IV: 1. The presence of positive cytology in CSF, 2. Positive cytology in pleural or abdominal fluids, 3. The presence of implants on pleural or peritoneal surfaces NOTE: The above excludes regional nodes and adjacent organ infiltration which places the patient in a more favorable grouping. NOTE: The presence of a pleural effusion or ascites without positive cytologic evaluation is not considered evidence of metastatic disease and the patient will not be considered to have Group IV disease.	Distant metastatic disease present at onset Although not limited to these, the following are considered evidence of metastatic disease: 1. The presence of positive cytology in CSF, 2. Positive cytology in pleural or abdominal fluids, 3. The presence of implants on pleural or peritoneal surfaces Note: Regional lymph node involvement and adjacent organ infiltration are not considered metastatic disease Note: The presence of a pleural effusion or ascites without positive cytologic evaluation is not considered evidence of metastatic disease

Evolution of the Clinical Group system

The definitions of the CGs are largely unchanged since its inception with the exception of minor alterations and clarifications. The changes are summarized in Fig. 1.

Figure 1. Evolution of the CG and COG modified TNM Classification Systems.

Utility and prognostic value of the Clinical Group system

In IRS I, treatment assignment was made based on CG alone. Since IRS II, additional prognostic factors have been used in conjuction with CG to determine treatment assignment.^6,7 CG has previously been shown to be prognostic.^3,4,6,7,28 Data from the IRS V/D-series and ARST series of studies confirms the prognostic value of CG (Fig. 2A). Of note, data from the IRS V/D-series and ARST series did not reveal significant differences in EFS or OS between CG II subgroups (Supplemental Document 1). However, CG II subgroup has impacted the radiation dose and has been incorporated into chemotherapy treatment assignment on certain COG protocols. Thus, treatment may modify the prognostic significance of CG II subgroups.^14,29

Figure 2: Outcome data.

A. EFS by clinical group; B. EFS by modified TNM stage; C. EFS by primary tumor site, including only patients with localized or regional disease; D. EFS by primary tumor size, including only patients with localized or regional disease; E. EFS by clinical regional lymph node status including only patients with localized or regional disease. F. Optimal primary tumor size cut off for EFS is shown as the point that maximizes the absolute value of log rank statistic (ABS(sk)). This is based on data in the COG database from the IRS-V / D-series studies D9602, D9802, D9803, and ARST studies ARST0331, ARST0431, ATRS0531, ARST08P1.

Changes to the Clinical Group system

The COG STS committee will make the following changes to the CG definitions, as shown in Table 1:

1. The CG I and III subgroups (i.e. group Ia, Ib, group IIIa and IIIb) will be discontinued. The IRS V/D-series and ARST series of studies did not collect data on these subgroups. They have not been used for treatment assignment and their significance is unclear.

2. CG I will be updated to remove recommendations regarding lymph node biopsy. Various factors must be considered in determining the role of pathological regional lymph node assessment.^30-37

3. In certain study protocols, CG II has been described as “Total gross resection with evidence of regional spread.” We will clarify the definition of CG II as localized disease with microscopic residual or grossly resected regional disease with or without microscopic residual.

Clarifications of Clinical Group definitions

The distinction between regional and distant lymph nodes is critical as regional lymph node disease involvement is considered CG II or III while distant lymph node metastases are considered CG IV. The regional lymph node basins by primary site are outlined in Supplemental Document 2. Additionally, it should be emphasized that if there is gross residual disease in involved regional lymph nodes, this is considered CG III because CG IIb does not allow for any residual disease and CG IIc allows for only microscopic residual disease.

Limitations of the Clinical Group system

As the IRSG experience evolved, it was recognized that there are limitations to the CG system.³⁸ First, the extent of surgical resection may be influenced by the expertise and decision making of the individual surgeon. Thus CG is not solely reflective of the biology of RMS. Second, while the extent of resection at diagnosis has been shown to have prognostic significance, the prognostic implications may not be uniform across all disease sites. For example, excellent disease control can be obtained in orbital sites with biopsy, radiation and chemotherapy without surgical resection.^10,39,40 Additionally, the IRSG and COG have emphasized retaining form/function for quality of life. Third, CG does not consider some factors that have been shown to be prognostic including primary tumor site, size, histology and molecular findings.^{3,6,7, 41} Finally, the IRSG recognized that their approach to classifying RMS differed from the cooperative groups in Europe, which limited the ability to compare treatments and outcomes.^38,42

COG RMS modified Tumor, Node, Metastasis (TNM) Staging System

Overview and history of the COG RMS modified TNM staging system

The recognition of the limitations of the CG system prompted the development of a modified TNM staging classification system.^2,38,41,42 Using the Pediatrique Société Internationale d'Oncologie Pédiatrique/International Society of Pediatric Oncology (SIOP)-UICC TNM staging system (Supplemental Document 3) as a framework, the IRSG evaluated factors for inclusion into a modified TNM staging system.⁴¹ This iterative process occurred during the IRS II – IRS IV pilot studies.

The IRSG analyzed the components of the SIOP-UICC TNM staging system – primary tumor invasiveness, clinical/radiographic regional lymph node status and clinical/radiographic metastatic status – and additional factors that had been found to be of prognostic importance - primary tumor site, size, and histology.^{2,3,6,7, 41} Initially, the IRSG retrospectively analyzed the prognostic significance of these factors using data from a subset of patients enrolled on IRS II from 1978-1982.⁴¹ In this analysis, primary tumor site, size, invasiveness, and metastatic status were each found to have prognostic significance. There was a strong relationship between tumor invasiveness and size and a tendency toward poorer survival for patients with tumors >5 cm regardless of tumor invasiveness. Regional nodal status and histology were not associated with overall survival in this analysis. This may have been due to the small number of patients with known regional nodal involvement and the lack of consensus on the definition of alveolar histology at this time as later studies have shown that both of these factors have prognostic significance.⁴¹ These results led to a prospective data collection on IRS III using a proposed modified TNM system which included primary tumor site and size, clinical/radiographic regional nodal status and clinical/radiographic metastatic status.² In contrast to the prior retrospective analysis, in IRS III, regional nodal status did show prognostic significance in this prospective analysis.² This modified TNM staging system was then first formally utilized for treatment assignment in IRS IV.

The COG RMS modified TNM system is considered a clinical staging system which classifies patients’ disease status into one of four stages, signified by Arabic numerals 1-4, based on physical examination and radiographic assessment prior to any therapeutic intervention. It considers the primary tumor site and size, status of regional lymph nodes and distant metastatic involvement. Tumor invasiveness is included in the system although, importantly, it does not alter stage assignment. Favorable sites include the orbit, other non-parameningeal head and neck locations, non-bladder-prostate genitourinary tumors, and liver/biliary tract while unfavorable sites include all other sites such as parameningeal, bladder/prostate, and extremities. The current system is shown in Table 2.

TABLE 2.

Current COG RMS modified TNM Pre-treatment Staging Classification

Stage	Sites	T	Size	N	M
1	Orbit Head and Neck (excluding parameningeal) Biliary Tract/Liver GU – non-bladder/non-prostate	T1 or T2	a or b	N0, N1 or Nx	M0
2	Cranial Parameningeal Bladder/Prostate Extremity Other (includes trunk, retroperitoneum, etc but excludes biliary tract/liver)	T1 or T2	a	N0 or Nx	M0
3	Cranial Parameningeal Bladder/Prostate Extremity Other (includes trunk, retroperitoneum, etc but excludes biliary tract/liver)	T1 or T2	a	N1	M0
			b	N0, N1 or Nx	M0
4	All	T1 or T2	a or b	N0, N1 or Nx	M1
Tumor • T1: confined to anatomic site of origin • T2: extension and/or fixative to surrounding tissue • a: ≤ 5 cm in diameter in size • b: > 5 cm in diameter in size Regional Nodes • N0: regional nodes not clinically involved • N1: regional nodes clinically involved as defined as 1) > 1 cm by CT or MRI or 2) 18FDG avid • Nx: clinical status of regional nodes unknown (especially sites that preclude lymph node evaluation) M: Metastases • M0: No distant metastases • M1: Distant metastases present • Note: The presence of positive cytology in pleural fluid, abdominal fluid, or CSF and the presence of implants on pleural or peritoneal are considered evidence of metastasis.

Evolution of the COG RMS modified TNM staging system

There have been few changes to the COG RMS modified TNM staging system since its inception. The changes that have been made are summarized in Fig. 1.

Notably, in D9602, the biliary tract was changed from an unfavorable site to a favorable site, based on the IRSG/COG experience to date during which there was low rate of death due to recurrent disease.⁴³ In ARST0331, the liver was also changed from an unfavorable to a favorable site and grouped with the biliary tract as RMS in the liver is believed to arise from intrahepatic bile ducts.

In the current open COG study, ARST1431, clinical regional lymph node involvement (N1) was defined as lymph nodes >1 cm by CT or MRI or lymph nodes with 18-FDG avidity. Similarily, RECIST 1.1 defines abnormal lymph nodes as those ≥ 1.0 cm (lymph nodes ≥1.5 cm are considered both abnormal and measurable in RECIST 1.1).^44-47 Current published data indicates that FDG avidity has a high specificity for regional lymph node involvement in RMS.^48-56 For example, lymph-node sampling and FDG-PET-CT was performed at diagnosis in 39 patients with non-metastatic RMS of the extremity enrolled on the European Paediatric Soft Tissue Sarcoma Group (EpSSG)-RMS2005 study and FDG-PET-CT had a specificity of 94.4% for regional lymph node involvement.⁴⁸ ARST1431 is collecting data to further analyze the concordance between lymph node size on cross sectional imaging, FDG avidity, and pathologic status and this may help clarify the role of FDG PET in determining N1 status.

Utility and prognostic value of the COG RMS modified TNM staging system

COG RMS modified TMN stage has been validated as a strong prognostic factor in RMS.⁵⁷ Data from the IRS V/D-series and ARST series of studies confirms the prognostic value of stage as shown in Fig. 2B. The prognostic value of stage and the individual factors considered in the staging system are also shown in Fig. 2C-E.

In the current COG RMS modified TNM staging system, primary tumor size is categorized as ≤ 5 cm or > 5 cm. The 5 cm transition point has previously been shown to be prognostic and was again found to be prognostic in our analysis as shown in Fig. 2D.⁵⁸ Data from the IRS V/D-series and ARST studies was further analyzed and showed that the optimal primary tumor size transition point was 4.2 cm for EFS and 5 cm for OS as shown in Fig. 2F and Supplemental Document 1.^24,59 Notably, given the broad age range and size of patients with RMS, it is not clear if primary tumor size should be considered in absolute terms.^60,61 Further investigation is warranted to consider the role of augmentation of tumor size by patient size in stage.

The primary tumor is classified as either locally non-invasive (T1 = primary tumor confined to organ or tissue of origin) or locally invasive (T2 = primary tumor extension and/or fixation to surrounding tissue) as defined by clinical examination and imaging.⁵ Tumor invasiveness is included in the COG RMS modified TNM staging system but it does not impact the stage assignment as it has not been considered independent of size.^2,41 However, tumor invasiveness has been shown to be prognostic and is utilized to guide a risk-based approach to staging evaluation.²² As shown in Supplemental Document 1, data from the IRS V/D-series and ARST studies confirmed significant association between tumor size (≤ 5 cm or > 5 cm) and invasive status (T1 or T2) (p<0.0001) with tumors ≤ 5cm more likely to be T1 (71%) compared to tumors >5cm (34%).

Change to the COG RMS modified TNM staging system

The COG STS committee will make the following changes to the definitions of modified TNM stage, as shown in Table 3:

TABLE 3.

Updated COG RMS modified TNM Pre-Treatment Staging Classification

Stage	Sites	Size	N	M
1	Orbit Head and Neck (excluding parameningeal) GU – non-bladder/non-prostate	a or b	N0, N1 or Nx	M0
2	Parameningeal Bladder/Prostate Extremity Other (includes trunk, retroperitoneum)	a	N0 or Nx	M0
3	Parameningeal Bladder/Prostate Extremity Other (includes trunk, retroperitoneum)	a	N1	M0
		b	N0, N1 or Nx	M0
4	All	a or b	N0, N1 or Nx	M1
Tumor • T(site)1: confined to anatomic site of origin • T(site)2: extension and/or fixative to surrounding tissue • a: ≤ 5 cm in longest diameter • b: > 5 cm in longest diameter Regional Nodes • N0: regional nodes not clinically involved • N1: regional nodes clinically involved as defined as ≥1 cm measured in short axis on CT or MRI • Nx: clinical status of regional nodes unknown (especially sites that preclude lymph node evaluation) M: Metastases • M0: No distant metastases • M1: Distant metastases present • Note: The presence of positive cytology in pleural fluid, abdominal fluid, or CSF and the presence of pleural or peritoneal implants are considered evidence of metastases.

1. The biliary tract will be reclassified as an unfavorable site given recent data revealing suboptimal outcomes (Figure 2C; Table 3).^62-65 Specifically, Aye et al. reported outcomes of the 17 patients with localized RMS of the biliary tract treated on most recent low-risk COG RMS studies, D9602 and ARST0331, with 5-year EFS of 70.6% and 5-year OS of 76.5%.⁶³

2. For clarity, the description of “cranial parameningeal” will be changed to “parameningeal.”

3. The size criteria of involved regional lymph nodes will be changed from “>1 cm” to “≥ 1 cm.” Clarification will be made that regional lymph node size is based on short axis measurement while primary tumor size is based on longest dimension measurement. These changes are consistent with guidance in prior COG protocols and RECIST 1.1.^44-47

4. 18-FDG avid will be removed as a criteria for N1 regional nodal status until further data regarding concordance with lymph node size and pathologic status are known.

Limitations of the COG RMS modified TNM staging system

The COG RMS modified TNM staging system relies on clinical examination and imaging findings for assessment of disease status. However, there may be occult disease involvement not detectable on exam or imaging and exam and imaging findings that are equivocal or discordant with pathology.

The degree of agreement between clinical and pathologic assessment of regional nodes has previously been shown to vary between primary sites.^30,32,33,37 Notably, boys >10 years of age with paratesticular tumors and patients with extremity tumors have a high incidence of regional nodal involvement and clinical assessment has been shown to underestimate this leading to site-specific recommendations regarding surgical regional lymph node staging.^31-37 The concordance of clinical and pathologic regional nodal status was analyzed across all sites using data from the IRS V/D-series and ARST studies, and these results are shown in detail in Supplemental Document 1. In brief, 37.6% (85 of 226) of patients with clinical N1 status were pathologically not involved (N0) and 4.8% (19 of 398) of patients with clinical N0 status were pathologically N1. Notably, these data pre-date the integration of size criteria and FDG avidity into the COG modified TMN staging system. Given this discordance, we advocate for pathologic assessment of clinically positive nodes, whenever feasible. This is in keeping with guidance in prior COG study protocols. Importantly, COG study protocols have required RT to the regional nodal basin for either clinical or pathological N1 status with the exception that clinically enlarged lymph nodes that are biopsied and pathologically N0 do not require irradiation.

The assessment of metastatic disease can also be challenging with current imaging modalities. For example, patients may have small pulmonary nodules that are indeterminate on imaging and are not amenable to biopsy. The European Paediatric Soft Tissue Sarcoma Group (EpSSG) compared patients treated on the EpSSG RMS 2005 protocol without pulmonary nodules and those with pulmonary nodules that were considered indeterminate (defined by ≤ 4 nodules < 5 mm or 1 nodule measuring ≥ 5 and < 10 mm), all of whom were classified as having localized disease.⁶⁶ There was no significant difference in the 5-year EFS or OS between these two groups. This supports not upstaging patients who have indeterminate pulmonary nodules as defined in this study and otherwise localized RMS.⁶⁶ Further investigation is needed to provide evidenced-based guidance for the radiographic asessessment of regional lymph node and metastatic disease.

Finally, as with CG system, the COG RMS modified TNM staging does not include certain factors which are known to have prognostic significance including histology, FOXO1 fusion status and age.¹³

Primary site assignment for COG RMS modified TNM staging

Accurate primary site assignment for RMS is critical as it impacts stage. Site is assigned by a multidisciplinary team based on clinical, imaging and surgical findings. An updated guide for primary site definitions is included in Supplemental Document 2.

European Approach to staging of RMS

There are currently 2 major European clinical trial groups for RMS: EpSSG and the Cooperative Weichteilsarkom Studiengruppe (CWS). ⁶⁷ Historically, the European cooperative groups used the SIOP-UICC TNM staging system (Supplemental Document 3).⁴² However, more recent studies have also used the CG classification system or a post-surgical pathologic TNM (pTNM) classification system (Supplemental Document 3) in combination with additional factors for risk stratification.^48,58,68,69 One of the key differences between the classification systems used by COG and European cooperative groups, is that primary site, which is included in the COG modified TNM staging system, is not included in the European staging systems and is, instead, integrated into risk stratification.

Adult approach to staging in soft tissue sarcomas

In adults, soft tissue sarcomas (STSs) are staged using AJCC STS TNM staging systems.⁷⁰ As opposed to the COG RMS modified TNM staging system, which is considered a clinical staging system, AJCC TNM stage integrates both clinical and pathological data. The AJCC Cancer Staging manual was last updated in 2017 as the 8^th edition and includes separate staging algorithms for STS based on the primary site of disease (head and neck; trunk and extremities; abdomen and thoracic visceral organs; retroperitoneum).^70-72 For all sites, assessment of the primary tumor and regional lymph node and distant metastatic involvement is recommended. As with the COG modified TNM staging system, primary site is considered. Size and tumor invasiveness criteria are dependent on the specific primary site. Additionally regional nodal involvement meets criteria for stage 4 disease for certain sites such as trunk and extremity. However, it has been proposed that this be reassessed given data showing distant metastatic disease is associated with poorer survival than regional nodal involvement.⁷¹

Conclusions

We have reviewed the systems used by COG to classify disease extent among patients with newly diagnosed RMS. Staging systems are critical to consistently and accurately classify extent of disease, to guide treatment decisions, and to enable comparisons of patient outcomes. To fulfill these goals, staging systems must be clear, valid, and uniformly adopted. Importantly, the approach to staging must be updated as we gain further knowledge about the biology of disease. However, this must be balanced with the consideration that consistency in staging systems across time enables accurate comparisons to historical cohorts.

CG and COG RMS modified TNM stage are complementary and, together, describe the anatomic extent of disease through clinical and pathologic evaluation. Data from recent COG studies confirm that CG and COG RMS modified TNM stage remain valuable prognostic tools. We made changes to these systems including removing the subgroups of CG I and III, re-classifying the biliary tract as an unfavorable site, and updating the guidance for what constitutes as clinically involved regional lymph nodes. As our understanding of RMS continues to evolve, it is relevant to continue to reassess and update our classifications.

The CG system and COG RMS modified TNM staging systems do not include certain factors now shown to be prognostic of outcome. Thus, starting with IRS V/D-series of studies, patients were assigned to treatment groups based on risk stratification, an approach that has continued to evolve based on increased knowledge gained through sequential studies. While a detailed discussion of the risk stratification system for RMS is beyond the scope of this article, data from IRS I through IRS IV provided the rationale for initiation of a formal risk stratification system.¹³ The current risk stratification utilizes CG, COG RMS modified TNM stage, histologic subtype and/or FOXO1 fusion status and age at diagnosis.^5,73-76 Future approaches may include additional molecular risk factors such as TP53, MYOD1, and MYCN amplification.⁷⁷

Although CG is used by COG and European organizations, there are differences in the TNM staging systems used and risk stratification approach which limit our ability to compare therapy and outcomes. We support ongoing international efforts through the International Soft Tissue Sarcoma Consortium (INSTRuCT) to share clinical data and harmonize approaches to classification of RMS.⁶⁷

The CG and COG RMS modified TNM stage remain important tools for classification of disease extent in RMS and must be used in combination with the risk stratification system to optimally guide prognosis and treatment decisions for patients.

Abbreviations Key:

CG

Clinical Group

COG

Children’s Oncology Group

CWS

Cooperative Weichteilsarkom Studiengruppe

EpSSG

The European Paediatric Soft Tissue Sarcoma Group

EFS

Event Free Survival

18-FDG

18-Fluorodeoxyglucose

IRSG

Intergroup Rhabdomyosarcoma Study Group

OS

Overall Survival

PET

Positron Emission Tomography

RMS

Rhabdomyosarcoma

TNM

Tumor Node Metastasis

SIOP

Pediatrique Société Internationale d'Oncologie Pédiatrique/International Society of Pediatric Oncology

UICC

Union for International Cancer Control