Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Comparative study of the POG and FNCLCC grading systems in non-rhabdomyosarcoma soft tissue sarcomas
(USC Thesis Other)
Comparative study of the POG and FNCLCC grading systems in non-rhabdomyosarcoma soft tissue sarcomas
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
COMPARATIVE STUDY OF THE POG AND FNCLCC GRADING SYSTEMS IN
NON-RHABDOMYOSARCOMA SOFT TISSUE SARCOMAS
by
Yan Yang
A Thesis Presented to the
FACULTY OF THE USC KECK SCHOOL OF MEDICINE
UNIVERSITY OF SOUTHERN CALIFORNIA
Copyright 2021
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF SCIENCE
(BIOSTATISTICS)
May 2021
Yan Yang
Dedication
To my family,
for their endless love and support.
ii
Acknowledgements
I would like to express my deepest gratitude to my advisor, Dr. Donald Barkauskas, for his
encouragement and guidance through the process of finishing this thesis. Thank you for intro
ducing me to ARST0332 study, guiding me to explore the wonderful world of survival analysis,
and helping me to learn l5f£X.
I also would like to thank my thesis committee members, Dr. Todd A Alonzo and Dr. Wendy
Jean Mack, for their valuable comments to improve this thesis significantly.
Special thanks to my dearest friend Elika Su for pushing me to finish this thesis.
Special thanks to my friends Huiyu Xie, Shuyao Ran, Yanyang Peng, and cat Neko for sup
porting me physically and emotionally during this pandemic time.
iii
Table of Contents
Dedication
Acknowledgements
List of Tables
List of Figures
Abstract
Chapter 1: Introduction
Chapter 2: Methods
2.1 Study Design and Participants
2.2 Histologic Grading
2.3 Treatment Outcome
2.4 Statistical Analysis
Chapter 3: Results
3.1 Histologic Grade and Treatment Outcome . . . . . .
3.2 Prognostic Factors for Discrepantly Graded Patients .
Chapter 4: Discussion
Tables
Figures
Bibliography
ii
iii
V
vi
vii
1
3
3
4
5
6
7
7
8
11
14
22
26
iv
List of Tables
1
2
3
4
5
Association between Tumor Grades and Event-free Survival .
Baseline Patient Characteristics by Regrouped Grade . . . . .
Pairwise Association between Variables among Discrepant Patients
Univariable Cox Proportional Hazards Model on Event-free and Overall
Survival among Discrepant Patients . . . . . . . . . . . . . . . . . . . . . . .
Final Multivariable Cox Proportional Hazards Models
14
15
18
19
21
V
List of Figures
1
2
3
4
Experimental Design Schema . . . . . . . . . . . . . . . . . . . . . . . . . . .
Event-free and Overall Survival of All Patients by POG Grade Stratified
by Risk Group . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Event-free and Overall Survival of All Patients by FNCLCC Grade
Stratified by Risk Group . . . . . . . . . . . . . . . . . . . . . . . . . . .
Event-free and Overall Survival of All Patients by Regrouped Grade
Stratified by Risk Group .................. ........ . .
22
23
24
25
vi
Abstract
A variety of histologic grading systems are available for soft tissue sarcomas, but no system
for pediatric non-rhabdomyosarcoma soft tissue sarcomas (NRSTS) has been consistently utilized
worldwide. The Pediatric Oncology Group (POG) and French Federation of Cancer Centers Sar
coma Group (FNCLCC) systems were assessed and compared to find which system is the best for
identifying pediatric NRSTS treatment outcomes. Utilizing the data from ARST0332, we exam
ined 528 eligible NRSTS patients whose tumors were graded by both POG and FNCLCC criteria.
After stratifying by risk group, both POG and FNCLCC systems were significantly associated
with event-free survival (EFS) (p = 0.04 and p = 0.005, respectively), but only FNCLCC grade was
significantly associated with overall survival (OS) (p = 0.03). Both grading systems showed in
creased ability when predicting EFS of low-risk patients. Grade discrepancies were observed in
15.5% of the study cohort. Multivariable analysis among discrepantly graded patients suggests
that risk group was a significant predictor for EFS; age, ethnicity, and tumor size were significant
predictors for OS. In conclusion, the FNCLCC grading system showed better predictive value than
the POG system for OS; both the POG and FNCLCC systems were prognostic factors for EFS, es
pecially useful for low-risk patients. For patients whose tumors received discrepant grades, risk
group appeared to be a key factor related to EFS; tumor size appeared to be a key factor related
to OS.
vii
Chapter 1
Introduction
Non-rhabdomyosarcoma soft tissue sarcomas (NRSTS) comprise approximately 3.5% of child
hood cancer cases [1]. Although the incidence rate is relatively rare, NRSTS collectively account
for 60% of all pediatric soft tissue sarcomas [2].
Histologic grading has demonstrated clinical importance for soft tissue sarcomas by its pre
dictive value for adult [3, 4] and pediatric patients [5]. Histologic grading is also a useful de
terminant for treatment choice. Several histologic grading systems have been developed for soft
tissue sarcomas. The French Federation of Cancer Centers (Federation Nationale des Centres de
Lutte Contre le Cancer; FNCLCC) system, proposed by Trojani et al., is one of the most utilized
systems for adult sarcomas [6]. This system has advantages in predicting local aggressive and
distant metastasis in adult patients [7]. The Pediatric Oncology Group (POG) grading system,
proposed by Parham et al., is a modification of the National Cancer Institute adult grading sys
tem [8]. The POG system accounts for unique clinical and morphologic entities in young children
[8]. Although both grading systems have demonstrated their predictive values in previous stud
ies, the FNCLCC system was designed to grade adult sarcomas but not for pediatric sarcomas
[6]; the POG system was proposed in 1995, some disease designations are currently outdated,
1
such as the term deep-seated dermatofibrosarcoma protuberans [9]. Currently, there is nearly no
published information assessing and comparing the two systems in pediatric NRSTS patients -
particularly patients treated prospectively in a standardized fashion.
ARST0332 is the first large scale prospective study for pediatric NRSTS that recruited more
than 500 patients in 159 hospitals from three countries. The novel risk-adapted treatment strat
egy from this study has proven to be effective for reducing mortality rates and limiting toxicity
[10]. During central pathology review, each tumor was evaluated by both the POG and FNCLCC
grading systems, which offered an opportunity to determine which grading system is the best for
future studies and examine mismatches in grading.
In this study, using data from ARST0332, we assess the prognostic utility of POG and FNCLCC
grading systems in pediatric NRSTS. Furthermore, we report clinical features and explore prog
nostic factors of discrepantly graded tumors to have a better understanding of the two grading
systems.
2
Chapter 2
Methods
2.1 Study Design and Participants
ARST0332 was open to accrual from 05 February 2007, to 10 February 2012. Patients aged
less than 30 years and newly diagnosed with NRSTS within 12 weeks were recruited for this
therapeutic study. On the basis of a novel risk classification system for pediatric NRSTS, eligible
patients were categorized into one of three risk groups and assigned to one of four treatment
arms.
Figure 1 displays the experimental design schema. The risk classification system was proposed
based on known prognostic factors, including maximal tumor diameter, metastasis status, POG
grade, and extent of tumor resection before enrollment [11-15]. The system has three distinct
categories [16]: The low-risk group included patients with non-metastatic, resectable low-grade
tumors, and those with non-metastatic, resectable high-grade tumors whose maximal diameters
were less than 5 cm. The intermediate-risk group included patients with non-metastatic, re
sectable high-grade tumors whose maximal diameters were greater than 5 cm or non-metastatic,
unresectable tumors. The high-risk group included all patients with metastatic tumors regardless
3
of other clinical features. Based on risk classification, patients were assigned to one of the four
following treatment arms [16]:
• Arm A: Observation
• Arm B: Adjuvant Radiotherapy
• Arm C: Adjuvant chemotherapy and radiotherapy
• Arm D: Neoadjuvant chemoradiotherapy
Treatment Arm A consisted of low-risk patients with non-metastatic, resectable low-grade
tumors or high-grade tumors whose maximal diameters were less than 5 cm and received nega
tive margins, together with those high-risk patients whose tumors received low grade and both
primary and metastatic tumors were resectable. Treatment Arm B consisted of low-risk patients
with non-metastatic, resectable high-grade tumors whose maximal diameters were less than 5
cm and received positive margins. Treatment Arm C consisted of intermediate-risk patients with
non-metastatic, resectable high-grade tumors whose maximal diameter were larger than 5 cm,
and those high-risk metastatic patients whose primary high-grade tumors were resectable. Treat
ment Arm D consisted of all patients with unresectable primary tumors.
2.2 Histologic Grading
According to the central pathology review, each tumor was graded by both POG and FNCLCC
grading systems. The POG grading system is a three-tier grading system devised on the basis of
the previous adult grading systems and accounted for the unique pathologic features for pedi
atrics [8]. Grade is determined by histologic type, percent of necrosis, mitotic count, and status
4
of nuclear atypia. It should be noted that the POG grade was utilized to determine the risk cate
gory and treatment arm at enrollment. The FNCLCC grading system is a widely used three-tier
grading systems for adult soft tissue sarcomas [6]. The grade is determined by summing scores
of tumor differentiation, mitotic count, and tumor necrosis status.
Based on characteristics and standard clinical practice of the two grading systems, we re
garded the POG grade 1 or 2 as low grade and POG grade 3 as high grade; the FNCLCC grade 1 as
low grade, and FNCLCC grade 2 or 3 as high grade. On the basis of these criteria, we regrouped
patients into three distinct categories as follows:
• Low grade: patients whose tumors were categorized as POG 1/2 and FNCLCC 1
• Discrepant: patients whose tumors were categorized discrepantly by POG and FNCLCC
(i.e., low in one system and high in the other)
• High grade: patients whose tumors were categorized as POG 3 and FNCLCC 2/3
2.3 Treatment Outcome
Our primary treatment outcomes were event-free survival (EFS) and overall survival (OS).
Event-free survival was defined as the measure of time from enrollment to the occurrence of
tumor relapse or progression, diagnosis of a second malignant neoplasm (SMN), death, or last
contact, whichever came first. EFS events were tumor relapse or progression, SMN, and death.
Overall survival was defined as the measure of time from enrollment to death or last contact,
whichever came first.
5
2.4 Statistical Analysis
To determine the relationship between histologic grading and clinical outcomes, log-rank
tests stratified by risk group were performed on OS and EFS outcomes for POG, FNCLCC, and
regrouped grade. Survival distributions were estimated using Kaplan Meier estimator and plotted
for visualization. The 5-year survival rates with 95% confidence interval (CI) were constructed
based on log-log transformation.
To investigate the prognostic factors for discrepantly graded patients, a secondary analysis
was performed. Patient baseline demographic and clinical features were compared by regrouped
grade. Demographic features included age, sex, race, and ethnicity. Clinical features included
primary tumor size, tumor depth, invasiveness, metastases, extent of resection, microscopic mar
gins, risk group and treatment arm. For continuous variables, one-way ANOVA was utilized and
Scheffe's Multiple Comparison tests were performed among significant variables. For categorical
variables, Fisher's Exact Test was performed. Then, within discrepantly graded patients, pairwise
Fisher's Exact tests were further used to examine the association among categorical variables.
Using Cox proportional hazards regression, each variable was assessed univariabely on both
EFS and OS after adjusting for risk group. Then, in multivariable analysis, backward selection
based on Akaike Information Criterion (AIC) was employed to select the best set of predictors.
Two-sided p-values were reported and the significance level was set as 0.05. All statistical
analyses were conducted in R (version 3.6.0), using the 'survival' and 'survminer' packages.
6
Chapter 3
Results
Between Feb 05, 2007, and Feb 10, 2012, a total of 588 patients were registered in ARST0332;
38 patients were excluded because of ineligibility. According to the protocol, 550 eligible patients
were assigned to one of the four risk-based treatment arms. For analysis, 21 patients were treated
differently than specified by the protocol, and one patient had an intermediate FNCLCC grade
tumor, leaving 528 patients in the analytic population. For the 528 patients, the estimated 5-year
EFS was 68.1% (95% CI: 64.1%, 72.4%), and the estimated 5-year OS was 79.6% (95% CI: 76.1%,
83.3%).
3.1 Histologic Grade and Treatment Outcome
All tumors were graded by both POG and FNCLCC criteria during central pathology review.
For the POG grading system, 60 tumors (11.4%) were evaluated as grade 1, 86 tumors (16.3%) as
grade 2, and 382 tumors (72.3%) as grade 3. For the FNCLCC grade system, 70 tumors (13.3%)
were evaluated as grade 1, 222 tumors (42.0%) as grade 2, and 236 tumors (44.7%) as grade 3.
As described in the Methods, we categorized POG 1/2 and FNCLCC 1 as low grade, POG 3 and
7
FNCLCC 2/3 as high grade, which results in matches and mismatches in tumor grading. Namely,
67 (12.7% ) patients were low grade in both systems, 379 (71.8% ) were high grade in both systems,
and 82 (15.5% ) patients were high grade in one and low grade in the other. Of the 82 discrepantly
graded patients, 79 were assigned to POG 1/2 and FNCLCC 3; 3 were assigned to POG 3 and
FNCLCC 1.
We evaluated the association between tumor grading and clinical outcomes, summarized in
Table 1. Stratified by risk group, log-rank tests indicated that, for POG, FNCLCC, and regrouped
grades, there were statistically significant associations between grades and EFS (POG: p = 0.04,
FNCLCC: p = 0.005, and regrouped: p = 0.007). However, these relationships did not persist in OS;
only FNCLCC grade was significantly correlated with OS (p = 0.03). As Figures 2 to 4 illustrate,
the differences in EFS mainly exist in low-risk strata and between the following groups: POG 1
and POG 3; FNCLCC 1 and FNCLCC 2/3; low grade and discrepant, high grade.
3.2 Prognostic Factors for Discrepantly Graded Patients
To investigate prognostic factors for patients with discrepant tumors, we first checked the
distribution of baseline characteristics for all patients by regrouped grades, whose results can
be seen in Table 2. We found statistically significant differences among three grade groups in
clinical features, including primary tumor size, tumor depth, invasiveness, metastasis status, ex
tent of tumor resection, risk group, and treatment arm. Specifically, tumors tend to be larger and
metastatic for high-grade tumors compared to low-grade and discrepant tumors, and were more
likely to be deep and invasive for the discrepant and high grade than the low grade. High-grade
8
patients were the most likely group to develop unresectable tumors, followed by discrepant pa
tients and by low-grade patients. However, we did not identify any difference in demographic
features among the tumor grades.
Within discrepant patients, we checked the pairwise association between all categorical vari
ables using Fisher's Exact tests; results are shown in Table 3. For demographic features, we found
that after ten years old, patients were significantly more likely to have larger tumors (p = 0.003);
after eighteen years old, patients were significantly more likely to develop unresectable tumors,
then be assigned to the intermediate-risk group and Arm C (p < 0.05). For clinical characteristics,
tumor size, depth, invasiveness, the extent of resection, and margins were pairwise associated (p
< 0.05), indicating that larger tumors are probably deep, invasive, and unresectable tumors; or re
sectable but still have positive microscopic margins after surgery. Uninformative significant pairs
were omitted here, for example, risk groups were partially determined by tumor size, metastases,
and the extent of resection, so naturally they are significantly associated.
Using Cox proportional hazard regression, we assessed the prognostic factors for EFS and OS
within discrepantly graded patients. However, as Table 4 shows, we failed to identify any signif
icant factor for either EFS or OS after adjusting for risk group in univariable analysis. Due to the
limited sample size with small numbers of events, we excluded the following variables: tumor
depth for both OS and EFS, and microscopic margins for OS. Specifically, no patient with super
ficial tumors experienced death or EFS events; one patient with positive margins and one patient
with negative margins have died. Additionally, due to collinearity, we excluded the following
variables: metastases, treatment arms, the extent of resection, whose information was already
included in the assignment of risk groups.
9
Further multivariable analysis for EFS demonstrated that risk group was significantly associ
ated with EFS (p = 0.002) after accounting for age, gender and invasiveness. Within discrepantly
graded patients, the risk of experiencing any EFS event in high-risk patients was 3.31 times (95%
CI: 0. 71, 15.45) that of intermediate-risk patients; in low-risk patients was 0.17 times (95% CI: 0.04,
0.68) compared to intermediate-risk patients. Furthermore, tumor size is a negative prognostic
factor for OS (p < 0.001) after accounting for age and ethnicity, but risk group was no longer
predictive of OS. Table 5 summarizes the results of multivariable analysis for EFS and OS.
Chapter 4
Discussion
Our study demonstrates that both POG and FNCLCC grading systems were significantly cor
related with EFS after stratifying by risk group; the FNCLCC system has better performance in
predicting tumor mortality. Both grading systems showed excellent ability in predicting EFS of
low-risk patients. For discrepantly graded patients, risk group is a significant predictor for EFS;
age, ethnicity, and tumor size are significant predictors for OS.
Due to inherent similarities, the fact that both the POG and FNCLCC systems have predictive
values for EFS is not surprising, which corresponds to the study by Khoury et al. [9]. Both systems
assign grades based on tumor histology, mitotic count, and necrosis percentage [6, 8], which are
proven prognostic factors for local recurrence, distant metastases, and overall survival of soft
tissue sarcomas [ 17 -20]. However, we also found that their predicting performance is restricted
to low-risk patients. No difference has been shown between grade and clinical outcomes when
assessing intermediate and high-risk patients, and further research is needed to investigate this
phenomenon.
The FNCLCC system has a better ability for predicting tumor mortality. This is in agreement
with previous studies that found that FNCLCC grade had a significant effect on overall survival
11
[7, 21]. We postulate that this increased ability of the FNCLCC system mainly stems from the
more reasonable grading process. The FNCLCC system places relatively equal weight on tumor
differentiation, mitotic count, and tumor necrosis; the POG system first examines the tumor his
tology to assign specific tumors to grade 1 or 3 regardless of mitotic count and necrosis. In view of
predictive values of mitotic count and tumor necrosis, ignoring these two factors at the beginning
of the grading assignment might not be a reasonable choice.
Older age might be a potential predictor of tumor mortality. Pairwise association tests among
discrepantly graded patients illustrated that patients were more likely to develop large tumors
after ten, and develop unresectable tumors after eighteen. The multivariable model of OS also
demonstrated an increased risk for patients older than eighteen.
Our study has limitations. The rareness of events and limited sample size of 82 discrepantly
graded patients are real challenges. In the univariable analysis, we excluded variables like tumor
depth and margins due to the rare treatment events. The estimated 95% confidence intervals of
significant factors are relatively wide in the final models. Furthermore, risk group was univari
abely significantly correlated with both OS and EFS, and the risk classification systems have been
validated in another study using the SEER database [22]. However, risk group did not present in
the final multi variable model of OS. Thus, we did not have confidence to draw definite conclusions
for prognostic predictors for patients whose tumors received discrepant grades.
Additionally, we need to be cautious about unknown values for variables. For example, eth
nicity was a significant predictor in the final model of OS, because the unknown group was differ
ent from Hispanic/Latino and Non-Hispanic/Non-Latino groups. After removing three patients
whose ethnicities were unknown, ethnicity was no longer a significant predictor for OS.
12
In conclusion, the FNCLCC grading system has better prognostic value for tumor mortality
than POG system. Both POG and FNCLCC systems are equally effective when predicting EFS,
especially useful for low-risk patients. Due to limited sample size, we cannot draw definite con
clusions on prognostic factors for grade discrepancy.
13
Tables
Table 1: Association between Tumor Grades and Event-free Survival
Risk POG N 5yrEFS FNCLCC N 5yrEFS Regrouped N 5yrEFS
Low 1 60 96.4% 1 69 98.4% Low 66 98.3%
2 65 91.2% 2 116 87.1% Discrepant 62 89.3%
3 97 82.8% 3 37 77.3% High 94 82.2%
Intermediate 1 0 1 1 100.0% Low 1 100.0%
2 17 54.5% 2 78 67.3% Discrepant 16 51.1 %
3 210 65.7% 3 148 63.4% High 210 65.7%
High 1 0 1 0 Low 0
2 4 25.0% 2 28 18.9% Discrepant 4 25.0%
3 75 21.3% 3 51 22.9% High 75 21.3%
Stratified p = 0.04 Stratified p = 0.005 Stratified p = 0.007
14
Table 2: Baseline Patient Characteristics by Regrouped Grade
Variables Discrepant Low Grade High Grade p-value
n = 82 n = 67 n = 379
Demographic Characteristics
Age at enrollment (years) 13.3 (5.5) 11.4 (5.2) 13.2 (5.6) 0.05
Age group at enrollment 0.31
0-9 19 (23.2%) 25 {37.3%) 102 (26.9%)
10-17 51 (62.2%) 35 (52.2 %) 214 (56.5%)
18+ 12 (14.6%) 7 (10.4 %) 63 (16.6%)
Gender 0.43
Male 36 (43.9%) 36 (53.7%) 173 (45.6%)
Female 46 (56.1%) 31 (46.3%) 206 (54.5%)
Race 0.53
White 62 {75.6%) 52 {77.6 %) 260 {68.6%)
Black or African American 8 (9.8%) 9 (13.4 %) 63 (16.6%)
Other 3 (3.7%) 1 (1.5%) 20 (5.3%)
Unknown 9 (11.0%) 5 (7.5%) 36 (9.5%)
Ethnicity 0.95
Hispanic or Latino 11 (13.4%) 9 (13.4%) 60 (15.8%)
Not Hispanic or Latino 68 (82.9%) 55 (82.1 %) 306 {80.7%)
Unknown 3 (3.7%) 3 (4.5%) 13 (3.4%)
15
Variables Discrepant Low Grade High Grade p-value
n= 82 n= 67 n = 379
Clinical Characteristics
Primary tumor size (cm) 5.5 (4.1) 4.8 (4.3) 8.9 (5.6) <0.001
Primary tumor size group (cm) <0.001
::;5 42 (51.2% ) 44 (65.7% ) 109 (28.8%)
> 5 40 (48.8% ) 23 (34.3% ) 270 (71.2%)
Primary tumor depth <0.001
Superficial 13 (15.9% ) 35 (52.2% ) 44 (11.6% )
Deep 69 (84.1% ) 32 (47.8% ) 335 (88.4%)
Risk group <0.001
Low 62 (75.6% ) 66 (98.5% ) 94 (24.8% )
Intermediate 16 (19.5% ) 1 (1.5% ) 210 (55.4%)
High 4 (4.9% ) 0 (0.0% ) 75 (19.8% )
Invasiveness <0.001
Non-invasive 40 (48.8% ) 50 (74.6% ) 139 (36.7%)
Invasive 42 (51.2% ) 17 (25.4% ) 240 (63.3%)
Metastases <0.001
Yes 4 (4.9% ) 0 (0.0% ) 75 (19.8% )
No 78 (95.1% ) 67 (100.0% ) 304 (80.2%)
Treatment arm <0.001
ArmA 61 (74.4% ) 66 (98.5% ) 78 (20.6% )
16
Variables Discrepant Low Grade High Grade p-value
ll= 82 ll= 67 n = 379
ArmB 1 (1.2%) 0 (0.0%) 16 (4.2%)
ArmC 0 {0%) 0 (0.0%) 111 (29.3%)
ArmD 20 {24.4%) 1 (1.5%) 174 (45.9%)
Microscopic margins 0.26
Positive 26 (41.9%) 30 (45.5%) 108 (52.7%)
Negative 36(58.1 %) 36 (54.5%) 97 (47.3%)
Inapplicable 20 1 174
Un-resected tumor <0.001
Yes 20 (24.4%) 1 (1.5%) 166 (43.8%)
No 62 {75.6%) 66 {98.5%) 213 (56.2%)
17
Table 3: Pairwise Association between Variables among Discrepant Patients
p-values
1
(2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)
(1) Age 0.329 0.908 0.604 0.003 0.148 0.066 0.608 <0.001 0.841 <0.001 0.002
(2) Gender 0.405 0.313 0.827 0.371 0.657 0.627 0.886 0.443 0.798 0.775
(3) Race 0.008 0.281 0.165 0.342 1.000 0.803 0.297 0.736 0.806
(4) Ethnicity 0.720 0.811 0.905 1.000 1.000 0.012 0.761 0.714
(5) Diameter 0.002 <0.001 0.052 <0.001 0.007 <0.001 <0.001
(6) Depth 0.006 1.000 0.036 0.005 0.031 0.069
(7) Invasive 0.116 0.020 0.020 0.020 0.023
(8) Metastases 0.003 * 0.003 <0.001
(9) Arm 0.419 <0.001 <0.001
(10) Margins * *
(11) Un-resected <0.001
(12) Risk
1
: p-values by Fisher's Exact test
*: Cannot be obtained because of zero event
18
Table 4: Univariable Cox Proportional Hazards Model on Event-free and Overall Sur-
vival among Discrepant Patients
EFS OS
Characteristics HR (95% CI) p-value HR (95% CI) p-value
Age at enrollment (years) 0.95 (0.85, 1.07) 0.39 1.16 (0.98, 1.36) 0.08
Age group at enrollment 0.26 0.10
0-9 1.0 1.0
10-17 0.30 (0.70, 1.26) 0.10 0.31 (0.02, 5.19) 0.42
18+ 0.38 (0.06, 2.36) 0.30 1.72 (0.09, 34.20) 0.72
Gender
Male 1.83 (0.66, 5.04) 0.24 1.35 (0.29, 6.38) 0.70
Female 1.0 1.0
Race 0.57 0.47
White 1.0 1.0
Black or African American 1.81 (0.39, 8.44) 0.45 3.02 (0.30, 30.13) 0.35
Other
*
0.99
*
0.99
Unknown 1.95 (0.52, 7.22) 0.32 4.10 (0.67, 25.25) 0.13
Ethnicity 0.37 0.30
Hispanic or Latino 1.0 1.0
Not Hispanic or Latino 0.41 (0.12, 1.33) 0.14 0.25 (0.04, 1.56) 0.14
Unknown 0.59 (0.06, 5.30) 0.64 0.86 (0.08, 9.66) 0.90
19
EFS OS
Characteristics HR (95% CI) p-value HR (95% CI) p-value
Tumor size (cm) 0.98 (0.81, 1.19) 0.83 1.22 (0.97, 1.54) 0.09
Tumor size group (cm)
::S:5 1.0 1.0
>5 0.90 (0.18, 4.44) 0.90 2.29 (0.14, 36.76) 0.56
Invasiveness
Non-invasive 0.46 (0.14, 1.56) 0.21 1.04 (0.16, 6.81) 0.97
Invasive 1.0 1.0
Microscopic margins
Positive 3.56 (0.69, 18.35) 0.13
Negative 1.00
*: Cannot be obtained because of zero event
20
Table 5: Final Multivariable Cox Proportional Hazards Models
Variable HR (95% CI) p-value
Event-free Survival
Age group 0.113
0-9 4.63 (1.05, 20.47)
10-17 1.0
18+ 1.87 (0.46, 7.50)
Gender 0.129
Female 1.0
Male 2.37 (0.77, 7.34)
Risk 0.002
Low 0.17 (0.04, 0.68)
Intermediate 1.0
High 3.31 (0.71, 15.45)
Invasiveness 0.097
Non-invasive 0.37 (0.11, 1.26)
Invasive 1.0
Overall Survival
Age 0.004
0-9 65.82 (1.14, 3803.13)
10-17 1.0
18+ 22.44 (2.55, 197.23)
Ethnicity 0.027
Hispanic or Latino 7.90 (0.93, 67.17)
Not Hispanic or Latino 1.0
Unknown 104.80 (2.46, 4460.83)
Tumor size (cm) 1.78 (1.26, 2.54) <0.001
Note. Model was chosen based on the Akaike's information criterion (AIC)
21
Figures
Figure 1: Experimental Design Schema
Diagnosis of eligible soft tissue sarcoma
~---- Non-metastatic----~
Grossly Resected
l
Tumor Grade
/ ~
Low High
l Maximal Tu! or Diameter
/ ~
Microscopic
margins
0or0
s5cm
/ ""'
Microscopict Microscopic
ma~ins ma~ins
Arm A
Obseivation
j
Arm B
Adjuvant
radiotherapy
>Scm
Arm C
Ad1uvant
chemo+ RT
Unresected *
(Unresectable
or
high grade
tumor > Scrn
where delayed
resection
planned)
Arm D
Neoad1uvant
chemorad10therapy
Metastatic
to lymph nodes and/or distant sites
l
Tumor Grade
/""'
Low High
All ,a J R " cted~ .
Yes No
l
Microscopic
margins
0or0
I
Arm A
Obseivation
Assess Resection Status
of Primary Tumor
/ ""
Grossly resected Unresected
j j
Arm C
Adjuvant
chemo+ RT
Arm D
Neoadjuvant
chemoradiotherapy
~ Low Risk~ ~ Intermediate Risk 7 1------- High Risk ---------t
t Negative microscopic margins defined as the presence of a cuff of non-malignant tissue measuring at least 5mm in all directions surrounding the tumor in the
operative specimen. When the tumor abuts fascia or periosteum and the fascia or periosteum is removed in continuity with the tumor specimen, this margin will
also be considered negative.
• See section 14.2.2 for guidelines regarding which tumors should remain unresected at study entry.
22
Figure 2: Event-free and Overall Survival of All Patients by POG Grade Stratified by Risk
Group
~
:c
ro
.D
e
a.
~
·1:
:,
(/)
Kaplan-Meier plot of Overall Survival by POG grade
With Number of Patients at Risk
-· · - POG 1, low Risk - POG 3, low Risk - POG 3, Intermediate Risk - POG 3, High Risk
POG 2, Low Risk - - POG 2, Intermediate Risk - - POG 2, High Risk
1.00
0.75
0.50
0.25
0.00
POG3, LowRisk
POG 2.lntermedlateRii k
POG3,lntermediateRisk
60
65
97
17
210
Stratified p = 0.133
59
62
93
17
195
54
59
89
16
POG 2,H,gh Rii k 3
3
45
52
83
11
166
2
Time
41 37
47 44
75 70
10 10
155 140
1
POG3. High Rlsk '--' 7.- 5 ___ _, 57'----3 ', 9----",-----",----=.--- 30 23 21
0 3
Time
27
36
58
10
112
16
8
18 14
23 16
37 15
8 7
68 43
11
8
Kaplan-Meier plot of Event-free Survival by POG grade
With Number of Patients at Risk
· · · · POG 1, Low Risk - POG 3, low Risk - POG 3, Intermediate Risk - POG 3, High Risk
POG 2, low Risk - - POG 2, Intermediate Risk - - POG 2, High Risk
1.00
0.75
0.50
0.25
Stratified p = 0.040"
0.00
3 5 8
Time
Number at risk
60 59 52 43 40 36 26 17 13
65 62 56 48 44 40 32 20 14
POG3, LowR isk 97 85 77 72 66 63 49 33 14
POG2.lntermed1 ateR1 ik 17 14 13 9 8 7 7 5 5
POG3.lntermedlate Rl$k 210 178 150 136 126 116 93 63 38
POG2. H,gh Risk 4 4 1 1 1 1 1
3
25
1
22
10
10
10
POG 3,Hog~ Ri$k L..C..... 7 ,' 5 ___ .c 3.- 3 ___ _, 20'----',-----','-----'i-----'i'-----.---.....;....;.----.---....;.--
17 14 13 10
0 5 8 10
Time
23
Figure 3: Event-free and Overall Survival of All Patients by FNCLCC Grade Stratified by
Risk Group
~
:c
ro
.D
e
a.
~
·1:
:,
(/)
Kaplan-Meier plot of Overall Survival by FNCLCC grade
With Number of Patients at Risk
· · -· FNCLCC 1, l ow Risk - FNCLCC 3, Low Risk - - FNCLCC 2, Intermediate Risk - - FNCLCC 2, High Risk
FNCLCC 2, Low Risk · -· · FNCLCC 1, Intermediate Risk - FNCLCC 3, Intermediate Risk - FNCLCC 3, High Risk
1.00
0.75
0.50
0.25 Stratified p = 0.034
0.00
4 5 6
Time
FNCLCC 1,Low Risk 69 66 60 53 50 48 35
FNCLCC2,Low Risl< 116 112 108 95 86 79 67
FNCLCC3.LowRisk 37 36 34 32 27 24 19
FNCLCC1.lnteffJIOOiateRisk 1 1 1 1 1 1 1
FNCLCC2, lnterroodiateRisk 78 76 70 63 57 55 45
FNCLCC3.lnterroodiatoRisk 148 135 123 113 107 94 76
FNCLCC2.HighRisk 28 24 16 11 7 7 5
FNCLCC3.HighRi sk '---' 5 .-1 ___ .; 37 c..... __ .-' 2 '," 6 ___ .;2""1 ---',------' 17 15 12
0
1.00
4 5 6
Time
Kaplan-Meier plot of Event-free Survival by FNCLCC grade
With Number of Patients at Risk
· · · · FNCLCC 1, Low Risk - FNCLCC 3, Low Risk - - FNCLCC 2, Intermediate Risk - - FNCLCC 2, High Risk
FNCLCC 2, Low Risk FNCLCC 1, Intermediate Risk - FNCLCC 3, Intermediate Risk - FNCLCC 3, High Risk
10
10
-----------------------------------
0.75
- ~ , - ' - - - - - 1 - - - - - - - - - - - - - - -
----------- -----------
0.50
0.25
Stratified p = 0.005•
0.00
0 4 5 6 7 8 10
Time
Number at risk
FNCLCC 1,Low Risk 69 66 59 52 49 47 34 24 17 4
FNCLCC2,Low Risk 116 108 98 85 78 71 57 39 20 6
FNCLCC3,LowRlsk 37 32 28 26 23 21 16 7 4 2
FNCLCC 1, lnt8ffll&diateR, si< 1 1 1 1 1 1 1 1 1 0
FNCLCC2, lnt~ iate Ri&k 78 67 60 52 46 44 39 28 20 9
FNCLCC3.lntermediateRisk 148 124 102 92 67 78 60 39 22 14
FNCLCC2,H,ghRisk 28 13 6 5 4 4 4 2 1 1
FNCLCC3.H,ghRisk 51 24 15 13 11 10 7 7 5 2
0 4 5 6 7 8 9 10
Time
24
Figure 4: Event-free and Overall Survival of All Patients by Regrouped Grade Stratified
by Risk Group
~
:c
ro
.D
e
a.
~
·1:
:,
(/)
1.00
0.75
0.50
0.25
0.00
Kaplan-Meier plot of Overall Survival by Regrouped grade
With Number of Patients at Risk
· --· Low grade, Low Risk - High grade, Low Risk - - Discrepancy, Intermediate Risk - - Discrepancy, High Risk
Discrepancy, Low Risk --· · Low grade, lnlermediate Risk - High grade, Intermediate Risk - High grade, High Risk
Stratified p = 0.093
3 4 5 6
Time
50 47 45 34 Low grade. Low Risk 66 63 57 24 17
50
"
39 30 Discrepancy. L ow Risk 62 61 59 18 14
80 72 67 57 High grade. Low Risk 94 90 86 36 14 4
1 1 1 1 Lowgrade,lntermediateRisk 1 1 1 1 1 0
10 9 9 9 O,scrapancy,lntermediateRisk 16 16 15 7 6 3
166 155 140 112 H igh grade.Intermediate Risk 210 195 178 68 43 25
2 1 1 1 D,screpancy. HighRisk 4 4 3 1 1 1
10
30 23 21 16 Highgr ado. HighRisk '---' 7 ,'. 5 ___ ", 57 ___ ..; 3 ',- 9---",-----','-----','---..c.;.----', 11'----.- 6---..; 2'----.---
1.00
3 4 5 6 8
Time
Kaplan-Meier plot of Event-free Survival by Regrouped Grade
With Number of Patients at Risk
· · · · Low grade, Low Risk - High grade, Low Risk - - Discrepancy, Intermediate Risk - - Discrepancy, High Risk
Discrepancy, Low Risk Low grade, Intermediate Risk - High grade, Intermediate Risk - High grade, High Risk
10
-------------------- -------------- ·
0.75
0.50
0.25
Stratified p = 0.007'
0.00
3 4 5 6 8 10
Time
Number at risk
L owgrade, L ow Risk 66 63 56 49 46 44 33 23 16 3
D,screpancy. L ow Risk 62 61 55 45 41 35 26 15 12 5
Highgr8de. L ow R 1$k 94 82 74 69 63 60 48 32 13 4
L.owgrade, lntermedl ateRisk 1 1 1 1 1 1 1 1 1 0
Dlscre?ancy, lntermediateRisk 16 13 12 8 7 6 6 4 4 1
High grade, lotermediateRjsk 210 178 150 136 126 116 93 63 38 22
();SC10?<1ncy,HighRisk 4 4 1 1 1 1 1 1 1 1
Highgr ade,High Risk '---' 7 ,'. 5 ___ ", 33'----' 2,'-0----',-----,'-----,'---..c.;.--- 17 14 13 10 8 5 2
1 2 4 5 6 7 8 9 10
Time
25
Bibliography
[1] Cynthia E. Herzog. "Overview of sarcomas in the adolescent and young adult
population". In: Journal of pediatric hematology/oncology 27.4 (2005), pp. 215-218.
[2] Sheri L. Spunt et al. "Non-Rhabdomyosarcoma Soft Tissue Sarcomas". In: Pediatric Bone
and Soft Tissue Sarcomas. Ed. by Alberto Pappo. Berlin, Heidelberg: Springer Berlin
Heidelberg, 2006, pp. 133-162.
[3] Andrea Ferrari et al. "Adult-type soft tissue sarcomas in pediatric-age patients:
Experience at the Istituto Nazionale Tumori in Milan". In: Journal of Clinical Oncology
23.18 (2005), pp. 4021-4030.
[ 4] Jean-Michel Coindre et al. "Predictive value of grade for metastasis development in the
main histologic types of adult soft tissue sarcomas". In: Cancer 91.10 (2001),
pp. 1914-1926.
[5] Sheri L Spunt et al. "Prognostic factors for children and adolescents with surgically
resected nonrhabdomyosarcoma soft tissue sarcoma: An analysis of 121 patients treated
at St Jude Children's Research Hospital". In: Journal of Clinical Oncology 17.12 (1999),
pp. 3697-3705.
[ 6] M. Trojani et al. "Soft-tissue sarcomas of adults; study of pathological prognostic
variables and definition of a histopathological grading system". In: International Journal
of Cancer 33.1 (1984), pp. 37-42.
[7] L Guillou et al. "Comparative study of the National Cancer Institute and French
Federation of Cancer Centers Sarcoma Group grading systems in a population of 410
adult patients with soft tissue sarcoma." In: Journal of Clinical Oncology 15.1 (1997).
PMID: 8996162, pp. 350-362.
[8] DM Parham et al. "Nonrhabdomyosarcomatous soft tissue sarcomas of childhood:
Formulation of a simplified system for grading". eng. In: Modern Pathology 8.7 (Sept.
1995), pp. 705-710.
26
[9] Joseph D. Khoury et al. "Grading of nonrhabdomyosarcoma soft tissue sarcoma in
children and adolescents: a comparison of parameters used for the Federation Nationale
des Centers de Lutte Contre le Cancer and Pediatric Oncology Group Systems." In:
Cancer 116.9 (2010), pp. 2266-2274.
[10] Sheri L Spunt et al. "A risk-based treatment strategy for non-rhabdomyosarcoma
soft-tissue sarcomas in patients younger than 30 years (ARST0332): a Children's
Oncology Group prospective study". In: The Lancet Oncology 21.1 (2020), pp. 145-161.
[11] Sheri L. Spunt et al. "Clinical Features and Outcome oflnitially Unresected Nonmetastatic
Pediatric Nonrhabdomyosarcoma Soft Tissue Sarcoma". In: Journal of Clinical Oncology
20.15 (2002). PMID: 12149295, pp. 3225-3235.
[12] Sheri L. Spunt et al. "Prognostic Factors for Children and Adolescents With Surgically
Resected Nonrhabdomyosarcoma Soft Tissue Sarcoma: An Analysis of 121 Patients
Treated at St Jude Children's Research Hospital". In: Journal of Clinical Oncology 17.12
(1999). PMID: 10577841, pp. 3697-3705.
[13] Alberto S. Pappo et al. "Metastatic nonrhabdomyosarcomatous soft-tissue sarcomas in
children and adolescents: The St. Jude Children's Research Hospital experience". In:
Medical and Pediatric Oncology 33.2 (1999), pp. 76-82.
[14] Karen C. Marcus et al. "Childhood soft tissue sarcoma: A 20-year experience". In: The
Journal of Pediatrics 131.4 (1997), pp. 603-607.
[15] JE McGrory et al. "Nonrhabdomyosarcoma soft tissue sarcomas in children. The Mayo
Clinic experience". In: Clinical orthopaedics and related research 374 (May 2000),
pp. 247-258.
[16] Sheri L Spunt et al. "Protocol of Risk-Based Treatment for Non-Rhabdomyosarcoma Soft
Tissue Sarcomas (NRSTS) in Patients Under 30 Years of Age. Children's Oncology Group
(ARST0332)". In: (2011).
[ 17] Hiroshi Hashimoto et al. "Prognostic significance of histologic parameters of soft tissue
sarcomas". In: Cancer 70.12 (1992), pp. 2816-2822.
[18] BG Kulander et al. "Grading of soft tissue sarcomas: necrosis as a determinant of
survival". In: Modern pathology: an official journal of the United States and Canadian
Academy of Pathology, Inc 2.3 (May 1989), pp. 205-208.
[19] A. M. Mandard et al. "Prognostic factors in soft tissue sarcomas. A multivariate analysis
of 109 cases". In: Cancer 63.7 (1989), pp. 1437-1451.
[20] J.A.M. van Unnik et al. "Grading of soft tissue sarcomas: Experience of the EORTC soft
tissue and bone sarcoma group". In: European Journal of Cancer 29.15 (1993),
pp. 2089-2093.
27
[21] Dario Callegaro et al. "Development and external validation of two nomograms to predict
overall survival and occurrence of distant metastases in adults after surgical resection of
localised soft-tissue sarcomas of the extremities: a retrospective analysis". In: The Lancet
Oncology 17.5 (2016), pp. 671-680.
[22] Timothy V. Waxweiler et al. "Non-Rhabdomyosarcoma Soft Tissue Sarcomas in Children:
A Surveillance, Epidemiology, and End Results Analysis Validating COG Risk
Stratifications". In: International Journal of Radiation Oncology*Biology*Physics 92.2
(2015), pp. 339-348.
28
Abstract (if available)
Abstract
A variety of histologic grading systems are available for soft tissue sarcomas, but no system for pediatric non-rhabdomyosarcoma soft tissue sarcomas (NRSTS) has been consistently utilized worldwide. The Pediatric Oncology Group (POG) and French Federation of Cancer Centers Sarcoma Group (FNCLCC) systems were assessed and compared to find which system is the best for identifying pediatric NRSTS treatment outcomes. Utilizing the data from ARST0332, we examined 528 eligible NRSTS patients whose tumors were graded by both POG and FNCLCC criteria. After stratifying by risk group, both POG and FNCLCC systems were significantly associated with event-free survival (EFS) (p = 0.04 and p = 0.005, respectively), but only FNCLCC grade was significantly associated with overall survival (OS) (p = 0.03). Both grading systems showed increased ability when predicting EFS of low-risk patients. Grade discrepancies were observed in 15.5% of the study cohort. Multivariable analysis among discrepantly graded patients suggests that risk group was a significant predictor for EFS
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Lymph node metastases in non-rhabdomyosarcoma soft tissue sarcoma
PDF
Infants in non-rhabdomyosarcoma soft tissue sarcoma
PDF
Body wall primary tumors in non-rhabdomyosarcoma soft tissue sarcoma
PDF
A novel risk-based treatment strategy evaluated in pediatric head and neck non-rhabdomyosarcoma soft tissue sarcomas (NRSTS) patients: a survival analysis from the Children's Oncology Group study...
PDF
Survival of children and adolescents with low-risk non-rhabdomyosarcoma soft tissue sarcomas (NRSTS) treated with surgery only: an analysis of 234 patients from the Children’s Oncology Group stud...
PDF
Extremity primary tumors in non-rhabdomyosarcoma soft tissue sarcoma: survival analysis
PDF
Risk factors and survival outcome in childhood alveolar soft part sarcoma among patients in the Children’s Oncology Group (COG) Phase 3 study ARST0332
PDF
Distant metastatic nonrhabdomyosarcoma soft tissue sarcomas in children and adolescents: clinical features and survival outcome among patients in Children's Oncology Group Phase 3 Study ARST0332
PDF
Effects of demographic and tumor characteristics on outcomes in children with malignant peripheral nerve sheath tumors
PDF
An analysis of disease-free survival and overall survival in inflammatory breast cancer
PDF
Randomized clinical trial generalizability and outcomes for children and adolescents with high-risk acute lymphoblastic leukemia
PDF
An assessment of necrosis grading in childhood osteosarcoma: the effect of initial treatment on prognostic significance
PDF
Differential methylation analysis of colon tissues
PDF
Comparison for efficacy and toxicity in 2 vs 3 cycles of cisplatin consolidation therapy in children nasopharyngeal carcinoma
PDF
An assessment of impact of early local progression on subsequent risk for the treatment failure in adolescent and young adult patients with non-metastatic osteosarcoma
PDF
Racial and ethnic disparities in delays of surgical treatment for breast cancer
PDF
Comparative study of laparoscopy vs. laparotomy for ovarian mass removal
PDF
Instability of heart rate and rating of perceived exertion during high-intensity interval training in breast cancer patients undergoing anthracycline chemotherapy
PDF
Racial/ethnic differences in patient experiences with health care in association with the Gleason’s score level at prostate cancer diagnosis: findings from the SEER-CAHPS data
PDF
Comparison of Cox regression and machine learning methods for survival analysis of prostate cancer
Asset Metadata
Creator
Yang, Yan
(author)
Core Title
Comparative study of the POG and FNCLCC grading systems in non-rhabdomyosarcoma soft tissue sarcomas
School
Keck School of Medicine
Degree
Master of Science
Degree Program
Biostatistics
Publication Date
04/14/2021
Defense Date
03/19/2021
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
non-rhabdomyosarcoma soft tissue sarcomas,OAI-PMH Harvest,pediatric cancer,survival analysis,tumor grading
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Barkauskas, Donald (
committee chair
), Alonzo, Todd (
committee member
), Mack, Wendy (
committee member
)
Creator Email
yanyang0w0@gmail.com,yyang667@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c89-441720
Unique identifier
UC11668089
Identifier
etd-YangYan-9454.pdf (filename),usctheses-c89-441720 (legacy record id)
Legacy Identifier
etd-YangYan-9454.pdf
Dmrecord
441720
Document Type
Thesis
Rights
Yang, Yan
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Tags
non-rhabdomyosarcoma soft tissue sarcomas
pediatric cancer
survival analysis
tumor grading