Pediatric Blood Cancer - 2024 - Nagpal - Ewing Sarcoma Among Children 5 Years of Age or Younger is It a Different

Download as pdf or txt
Download as pdf or txt
You are on page 1of 9

Received: 29 November 2023 Revised: 19 July 2024 Accepted: 30 July 2024

DOI: 10.1002/pbc.31268 Pediatric


Blood &
Cancer The American Society of
Pediatric Hematology/Oncology
RESEARCH ARTICLE

Ewing sarcoma among children 5 years of age or younger: Is it a


different disease?

Chitrakshi Nagpal1 Shuvadeep Ganguly2 Archana Sasi1 Vivek Kumar3


Bivas Biswas4 Deepam Pushpam1 Akash Kumar5 Sandeep Agarwala6
Vishesh Jain6 Anjan Dhua6 Devender Kumar Yadav6 Shah Alam Khan7
Adarsh Barwad8 Asit Ranjan Mirdha8 Ahitagni Biswas9 Sanjay Thulkar10
Sameer Bakhshi1
1
Department of Medical Oncology, Dr. B.R.A. IRCH, All India Institute of Medical Sciences, New Delhi, India
2
Department of Medical Oncology, Jawaharlal Institute of Post Graduate Medical Education and Research, Puducherry, India
3
Division of Neonatology, Department of Paediatrics, All India Institute of Medical Sciences, New Delhi, India
4
Department of Medical Oncology, Apollo Multispecialty Hospital, Kolkata, India
5
Department of Medical Oncology, National Cancer Institute, All India Institute of Medical Sciences, Jhajjar, India
6
Department of Pediatric Surgery, All India Institute of Medical Sciences, New Delhi, India
7
Department of Orthopedics, All India Institute of Medical Sciences, New Delhi, India
8
Department of Pathology, All India Institute of Medical Sciences, New Delhi, India
9
Department of Radiation Oncology, Dr. B.R.A. IRCH, All India Institute of Medical Sciences, New Delhi, India
10
Department of Radiodiagnosis, Dr. B.R.A. IRCH, All India Institute of Medical Sciences, New Delhi, India

Correspondence
Sameer Bakhshi, Department of Medical Abstract
Oncology, Dr. B.R.A. Institute Rotary Cancer
Hospital, All India Institute of Medical
Introduction: Children ≤5 years of age with Ewing’s sarcoma (ES) possibly have a
Sciences, New Delhi 110029, India. distinct disease biology, data on which are scarce. We evaluated clinical features,
Email: sambakh@hotmail.com
outcomes, and prognostic factors of ES among children with age ≤5 years.
Shuvadeep Ganguly, Department of Medical
Oncology, Jawaharlal Institute of Post
Methods: Children with ES registered between 2003 and 2019 were included. Baseline
Graduate Medical Education and Research, clinical and treatment details were retrieved from medical records. Prognostic fac-
Puducherry 605006, India.
Email: ganguly.shuvadeep@gmail.com
tors were identified using multivariable Cox regression. Clinical features and outcomes
of children ≤5 years were compared with those greater than 5 years by chi-square
and log-rank tests. Propensity score-matched (PSM) analysis was done to evaluate the
impact of age on survival in the metastatic and localized subgroups.
Results: Out of the 859 patients, 86 (10%) were ≤5 years of age (median age 4 years,
60 males [69.8%]). The most common location was the extremities (37.2%), followed by
thorax (27.9%) and head and neck (H&N) (22.1%); baseline metastases were seen in 25
patients (29.8%). The median event-free-survival (EFS) and overall survival (OS) were
25.6 and 68.7 months, respectively. Metastatic disease predicted inferior OS (hazard

Abbreviations: CI, confidence interval; CT, computed tomography; EFS, event-free survival; EICESS, European Intergroup Cooperative Ewing’s Sarcoma Study; ES, Ewing sarcoma;
EWSR1:NFATC2, Ewings sarcoma RNA binding protein 1–nuclear factor of activated T cells 2; HR, hazard ratio; IQR, interquartile range; OS, overall survival; VAC/IE, vincristine, adriamycin,
cyclophosphamide, ifosfamide, etoposide.

Pediatr Blood Cancer. 2024;e31268. wileyonlinelibrary.com/journal/pbc © 2024 Wiley Periodicals LLC. 1 of 9


https://doi.org/10.1002/pbc.31268
15455017, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/pbc.31268 by Hacettepe Universitesi, Wiley Online Library on [16/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
2 of 9 NAGPAL ET AL .

ratio [HR] = 2.54, p = .018) and EFS (HR = 2.47, p = .007], symptom duration ≤3 months
predicted an inferior OS (HR = 2.17, p = .048). Compared to age greater than 5 years,
younger children had more H&N and less pelvic primaries (p < .001) and lesser baseline
metastases (p = .037). PSM analysis did not reveal any significant impact of age on OS
in the metastatic (HR = 1.59, p = .29) or localized cohort (HR = 1.77, p = .09).
Conclusions: Children with ES ≤5 years of age have a distinct favorable clinical pre-
sentation. However, age is not an independent prognostic factor for survival outcomes
when adjusted for confounders.

KEYWORDS
Ewing sarcoma, less than five, pediatric, primitive neuroectodermal tumor, prognostic factors

1 INTRODUCTION 5 years or younger to further the understanding of disease behavior in


this age group.
Ewing sarcoma (ES) is the second most common bone cancer in
children.1 It is primarily a cancer of children and adolescents, and
it accounts for 10%–15% of all primary bone tumors in this group. 2 MATERIALS AND METHODS
Patients with ES have a median age between 14 and 15 years.2,3
It is rarely reported in children younger than 5 years with an inci- 2.1 Study design
dence of less than 10 cases per million population per year.4,5
Thus, there are very limited data available on the clinical fea- This is a retrospective study that included patients with ES registered
tures and treatment outcomes in this population. It has been noted at a tertiary cancer care center in India between January 2003 till
that ES in younger patients may have a distinct phenotype, with December 2019. All patients with histologically confirmed diagnosis
higher incidence of extraosseous tumors, more head and neck, and of ES, who were treatment naïve, were eligible for the study. The data
lesser pelvic primaries, smaller tumor volumes, and a lower inci- collected from patient records included the date of diagnosis, baseline
dence of metastatic disease at presentation.5,6 However, small sam- clinical features including age, gender, presence of systemic symp-
ple sizes and heterogeneity in the differences noted across stud- toms, duration of symptoms prior to presentation, tumor diameter,
ies have not allowed for any firm conclusions to be drawn. Fur- tumor site, presence of upfront metastatic disease, and baseline rou-
thermore, the biological rationale for such differences is not well tine laboratory investigations. Patients with incomplete clinical details
understood.4,7–10 or those who received treatment outside were excluded. The study was
In ES across age groups, clinical and laboratory parameters at approved by institute ethics committee (IEC-594/06.06.2021). The
baseline such as metastatic disease at presentation, tumor volume need for informed consent was waived off due to retrospective nature
more than 100 mL, tumor size more than 8 cm, axial primaries, of the study.
and elevated levels of serum lactate dehydrogenase (LDH) have
been described to be adverse prognostic factors for survival.2,3,7,8,11
However, the clinical relevance of age as a prognostic factor is 2.2 Diagnostic and staging workup
controversial.6,12–14 Patients with localized ES of age less than
15 years had a better relapse-free survival in the European Inter- The diagnosis was confirmed by characteristic histopathological fea-
group Cooperative Ewing’s Sarcoma Study (EICESS) analysis; however, tures seen on biopsy. ES was diagnosed if the biopsy specimen revealed
age was not found to be prognostic of survival in the EW88 study the presence of small blue round cell tumor on morphology, and positiv-
in patients with localized disease.3,7 It is not known whether prog- ity for CD99 (MIC 2) and/or NKX2.2 on immunohistochemistry. Other
nostic factors for treatment outcomes are different in children aged round cell tumors were differentiated from ES by common leukocyte
≤5 years. antigen, desmin and myogenin immunohistochemistry. Florescent in
There is no prior patient dataset that has examined disease behav- situ hybridization (FISH) and/or molecular testing for EWSR1 (Ewings
ior of ES in children belonging to the age group of ≤5 years. Differences sarcoma RNA binding protein 1) translocation were not routinely avail-
in clinical presentations and/or outcomes may imply a difference in able during the study period, and hence the same was done only if
disease biology, for which tailoring of therapy may help in improving there was a diagnostic dilemma. The patients were then discussed
outcomes. In this study, we aimed to evaluate the clinical characteris- in a multidisciplinary team after which all decisions were taken both
tics, outcomes, and prognostic factors of ES in a cohort of children aged for diagnosis and treatment. Staging investigations included local site
15455017, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/pbc.31268 by Hacettepe Universitesi, Wiley Online Library on [16/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
NAGPAL ET AL . 3 of 9

imaging with computed tomography (CT) or magnetic resonance imag- albumin (<3.5 g/dL) were included as potential prognostic factors
ing (MRI). Metastatic workup included bone marrow aspiration and for survival in patients aged ≤5 years. Their impact on survival was
biopsy, CT scan of the chest with a bone scan, or whole-body 18 F- assessed using univariable Cox regression analysis. Multivariable Cox
fluorodeoxyglucose-positron emission tomography-computed tomog- regression analysis was performed in a forward stepwise manner on
raphy (18 F-FDG-PET/CT) scan in all patients. the factors identified in univariable analysis with p-value ≤.1. Factors
with p-value less than .05 in this model were considered to have a
significant prognostic impact on survival.
2.3 Treatment administered

Standard first-line treatment was given in localized disease. Patients 2.5.2 Survival analysis
with localized disease uniformly received chemotherapy with vin-
cristine, doxorubicin, cyclophosphamide (VAC), alternating with Kaplan–Meier analysis was performed using survival17 and survminer18
ifosfamide-etoposide (IE). Initially it was given at a 3-weekly interval; packages in R. Survival data were reported in the form of median EFS
however, after 2017,15 it was given in a dose-dense 2 weekly interval, or OS with 95% confidence intervals (CI). The survival outcomes of
followed by local control with either surgery or radiotherapy, and then children ≤5 years were compared with those older than 5 years by
adjuvant therapy with the same drugs. Doxorubicin was replaced with Cox regression analyses. Listwise deletion was used for missing vari-
D-actinomycin when its cumulative dose was more than 375 mg/m2 . ables. To account for the imbalance of confounders between the two
In patients with metastatic disease, VAC/IE chemotherapy was age groups (age ≤5 vs. >5 years), a propensity score matched anal-
given in 3 weekly cycles. In these patients, those who had a good ysis was done using matchit package in R.19 A propensity score was
response to neoadjuvant chemotherapy (NACT) at the primary as well estimated for each patient using a logistic regression model with all
as metastatic site(s), they were also administered local therapy in relevant covariates—sex, presence of metastases, tumor site, tumor
the form of surgery or radiation. Adjuvant chemotherapy was con- size, symptom duration, presence of fever, hemoglobin, total leukocyte
tinued following local therapy. The decision regarding the choice of count, and albumin in the metastatic and localized cohort separately
surgery and/or radiation was taken after discussion in multidisciplinary to identify the impact of age on the survival outcomes. Matched pairs
tumor board. After 2017, the chemotherapy in metastatic disease was were developed between the two age groups (≤5 vs. >5 years) in 1:4
changed to a VAC-only regimen without the use of IE.16 ratio by optimal matching without replacement, and the balance of
matched cohort was analyzed using standardized mean difference and

2.4 Outcomes love plot. Finally, the effect of age group in the propensity matched
cohort on EFS and OS was analyzed by multivariable Cox regression

The primary outcomes assessed were event-free survival (EFS), analysis.

defined as the time from date of diagnosis to the time of disease recur-
rence, progression or death, and overall survival (OS) defined as the
time between date of diagnosis till death due to any cause. The data 3 RESULTS
were censored on December 31, 2022.
3.1 Baseline characteristics

2.5 Statistical analysis During the study period from 2003 to 2019, a total of 1058 patients
with histologically confirmed, treatment-naive ES were registered at
Data analysis was done using Stata (Stata Statistical Software: Release our center. Out of the 856 evaluable patients, 86 (10%) were ≤5 years
15, StataCorp) and R version 4.3.0 (R Project for Statistical Computing). of age (median age 4 years, 60 males [69.8%]) (Figure 1). The baseline
Descriptive statistics were used to describe demographic details and demographic and clinical characteristics of the patients (for both,
baseline clinical features. Categorical data were expressed in the form those with age ≤5 and >5 years) are shown in Table 1. The median
of percentages, and continuous variables as median with interquartile duration of symptoms prior to presentation in patients ≤5 years of

range (IQR). Intergroup comparisons between age more than 5 and age was 3 months (IQR: 2–6 months). Twenty-three children (26.7%)

≤5 years were performed using chi-square tests and unpaired t-test. had extra-osseous primary tumors, and baseline metastases was
observed in 25 patients (29.8%). The most common primary site was
the extremities (37.2%), followed by the thorax (27.9%) and then the

2.5.1 Identification of prognostic features head and neck (22.1%). Compared to age more than 5 years, younger
children had a higher proportion of tumors for which head and neck

Clinical characteristics such as gender, duration of symptoms, sys- was the primary site (22.1% vs. 9.8%) and less pelvic primary tumors

temic symptoms at baseline, tumor diameter, primary tumor site (head (4.6% vs. 18.4%), shorter duration to symptoms prior to presentation

and neck, thorax, abdomen, pelvis, spine, and extremities), low blood (3 vs. 4 months), smaller tumor sizes (7 vs. 9 cm), and lower prevalence

hemoglobin (<10 g/dL), platelet count (<150 × 103 /μL), and serum of metastases at baseline (29.8% vs. 41.5%) (Table 1). There was no
15455017, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/pbc.31268 by Hacettepe Universitesi, Wiley Online Library on [16/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
4 of 9 NAGPAL ET AL .

FIGURE 1 Workflow of the study.

difference in the types of local therapy received in the form of either (68.7 months [25.8 not reached], 24.5 [12.7 not reached] and 30.3 [13.3
surgery, radiotherapy, or both between the younger cohort ≤5 years not reached], p = .06]. The 3-year OS and EFS in the localized subgroup
and the older one (Table 1). of the patients ≤5 years versus those more than 5 years was 66.7%
(49.9%−79.0%) versus 56.9% (51.6%−61.2%) (p = .20) and 54.8%
(38.8%−68.3%) versus 47.1% (42.0%–52.1%) (p = .40) (Figure 2A,B).
3.2 Prognostic factors for OS and EFS The 3 years OS and EFS in the metastatic subgroup of the patients
≤5 years versus those older than 5 years was 33% (11.7%−56.9%) ver-
On univariable analysis, total leukocyte count greater than sus 23.5% (18.1%−29.4%) (p = .24) and 28.0% (10.6%−48.6%) versus
11,000/mm3 , symptom duration less than 3 months, and metastatic 16.8% (12.2%−22.0%) (p = .18) (Figure 3A,B).
disease at baseline were found to be predictive of inferior OS. Among Propensity score-matched analysis in the metastatic cohort
these, symptom duration less than 3 months (hazard ratio [HR] = 2.17 revealed no difference in the OS between patients who were ≤5 years
[1.01–4.54], p = .048) and metastatic disease at baseline (HR 2.54 compared to the older cohort (HR = 2.15, 95% CI: 0.86–5.35, p = .105),
[1.18–5.49]; p = .018) were independent predictors of OS in mul- as well as in the localized cohort (HR = 1.74, 95% CI: 0.89–3.39,
tivariable analysis (Table 2). Metastases at baseline and symptom p = .100). Similarly, there was no difference in the EFS on propensity
duration less than 3 months were associated with an inferior EFS on score matching in the metastatic (HR = 1.39, 95% CI: 0.62–3.10,
univariable analysis, while only metastatic disease at baseline (HR p = .42) as well as the nonmetastatic cohort (HR = 1.35, 95% CI:
2.47 [1.29–4.76], p = .007) was an independent predictor of EFS on 0.77–2.36, p = .30).
multivariable analysis (Table 3). Patients with lung-only metastasis
had a better survival in the overall cohort (23.9 vs. 15.6 months,
p = .03), but not in the younger cohort (22.1 vs. 26.8 months, p = .95). 4 DISCUSSION
Tumor necrosis data were available in a total of 110 patients in
the overall cohort: the median OS seen in those with necrosis less In this study, we analyzed patients ≤5 years of age with ES treated
than 90% (n = 70) was 37.6 months (18.4 not reached) versus not at our center with a uniform protocol over a period of 16 years. We
reached (27.4 not reached) in those with necrosis ≥90% (p = .04). found that young children with ES had a higher prevalence of head and
In the younger cohort, necrosis data were available in 10 patients, neck and lesser pelvic origin tumors. This cohort tended to present
with 100% necrosis seen in six, 70% in one and no necrosis in three earlier, with smaller tumors at presentation and lower proportion of
patients. patients with baseline metastatic disease than the older cohort. These
findings are congruent with the results of various studies, including
Ioris et al’s cohort of children less than 6 years of age with primary
3.3 Impact of age on survival outcomes bony ES.4–6,13,20 The lower prevalence of metastasis may be attributed
to more head and neck and lesser pelvic primary sites, or to inher-
In the overall cohort, in the age group ≤5 versus 6−10 and more than ent differences in the tumor biology. Unlike head and neck tumors,
10 years, there was no significant difference in EFS (25.6 months [16.4– pelvic origin tumors are more frequently metastatic at presentation
68.7], 19.9 months [17.5–25.0] and 21.6 [19.3–25.9], p = .38) and OS and spread into the visceral space prior to symptom onset.20–22 The
15455017, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/pbc.31268 by Hacettepe Universitesi, Wiley Online Library on [16/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
NAGPAL ET AL . 5 of 9

TA B L E 1 Comparison of baseline demographic, clinical characteristics, and treatment administered to Ewing sarcoma patients based on age
group.

Age ≤5 years Age >5 years


(n = 86) (n = 773) p-value
Age (in years) 4 (3–5) 16 (12–21)
Median (IQR)
Sex, n (%)
Males 60 (69.8%) 537 (69.5%) .96a
Females 26 (30.2%) 236 (30.5%)
Fever at presentation, n (%) 28 (33.3%) 182 (23.6%) .054a
n = 84 n = 772
Symptom duration (months) 3 (2–6) 4 (3-8) .0002b
Median (IQR) n = 83 n = 770
Tumor diameter (cm) 7 (5–10) 9 (6–12) <.0001b
Median (IQR) n = 70 n = 624
Tumor site, n (%)
Skeletal 63 (73.2%) 687 (88.9%) .071a
Axial 36 (57.1%) 242 (35.2%)
Appendicular 27 (42.9%) 445 (64.8%)
Extra-skeletal 23 (26.7%) 86 (11.1%) .22a
Overall tumor site, n (%) <.001a
Extremities 32 (37.2%) 307 (39.7%)
Thorax 24 (27.9%) 149 (19.3%)
Head and neck 19 (22.1%) 76 (9.8%)
Pelvis 4 (4.6%) 142 (18.4%)
Spine 4 (4.6%) 58 (7.5%)
Abdomen 3 (3.5%) 41 (5.3%)
Metastatic disease at baseline 25 (29.8%) 321 (41.5%) .037a
n = 84c n = 773
Lung 15 (60%) 200 (62.5%) .82a
Bone marrow 4 (16%) 71 (22%) .48a
Bone 10 (40%) 151 (47%) .50a
Lung-only metastasis 10 (40%) 120 (37.4%) .80a
Hemoglobin (g/dL) 10.6 (9.3–11.4) 11.3 (9.9–12.8) .0002b
Median (IQR) (n) (n = 80) (n = 746)
TLC (per mm3 ) 9300 (7340–12,395) 8770 (6900–11,300) .15b
Median (IQR) (n) (n = 80) (n = 743)
Albumin (g/dL) 4.3 (4–4.6) 4.3 (3.9–4.7) .93b
Median (IQR) (n) (n = 70) (n = 688)
Local treatment, n (%) 57 (71.2%) 501 (67.5%) .94a
(n = 81) (n = 742)
Surgery, n (%) 19 (23.4%) 168 (22.6%)
RT, n (%) 26 (32.1%) 218 (29.4%)
Both surgery and RT, n (%) 12 (14.8%) 115 (15.5%)

Note: Column 1: median (IQR) or n (%).


Abbreviations: cm, centimeter; g/dL, grams per deciliter; IQR, interquartile range; mm3 , cubic millimeter; RT, radiotherapy; TLC, total leukocyte count.
a
Chi-square test.
b
Wilcoxon signed rank test.
c
Data for 2 patients not available.
15455017, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/pbc.31268 by Hacettepe Universitesi, Wiley Online Library on [16/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
6 of 9 NAGPAL ET AL .

TA B L E 2 Univariable and multivariable analysis of factors predictive of overall survival in patients of age ≤5 years.

Univariable analysis Multivariable analysis


Factor Category Median (IQR) OS HR 95% CI p-Value HR 95% CI p-Value
Sex Female (n = 26) NR (13.3–) 0.82 0.38–2.13 .82
Male (n = 60) 50.7 (22.1–NR) 1
Fever at onset Yes (n = 28) NR (13.3–) 0.91 0.38–2.14 .83
No (n = 56) 50.7 (25.2–NR) 1
Tumor diameter (cm) ≤5 (n = 21) NR (25.8–NR) 1 .41
>5 (n = 49) 68.7 (22.3–NR) 1.89 0.68–5.3 .20
Site of disease Head and neck (n = 19) NR (9.1–) 1
Extremity (n = 32) NR (15.8–) 1.11 0.39–3.13 .84
Thorax (n = 24) NR (22.1–) 0.77 0.25–2.39 .65
Abdomen (n = 3) 22.3 (22.3–NR) 0.91 0.11–7.59 .93
Pelvis (n = 4) 20.6 (16.4–NR) 2.41 0.59–9.81 .22
Spine (n = 4) 26.8 (13–NR) 2.33 0.572–9.27 .23
Metastasis at baseline Yes (n = 24) 25.2 (16.4–37.9) 2.22 1.04–4.71 .038 2.54 1.18–5.49 .018
No (n = 60) NR (25.8–) 1
Hemoglobin (g/dL) <10 (n = 52) 68.7 (25.17–NR)
≥10 (n = 28) 26.8 (13.3–NR) 1.31 0.58–2.92 .52
3
TLC (×10 /μL) >11 (n = 29) 37.9 (12.5–NR) 1.72 0.81–3.69 .16
≤11 (n = 51) NR (25.17–NR)
Symptom duration (months) ≥3 (n = 46) NR (25.8–NR) 1
<3 (n = 37) 37.9 (13.3–NR) 1.85 1.77–4.0 .10 2.17 1.01–4.54 .048
Tumor origin Skeletal (n = 63) 37.9 (22.1–NR)
Extra-skeletal (n = 23) NR (13.3–NR) 0.80 0.34–1.88 .60

Note: Column 3: median (IQR).


Abbreviations: CI, confidence interval; cm, centimeters; g/dL, grams per deciliter; HR, hazard ratio; IQR, interquartile range; NR, not reached; OS, overall
survival; TLC, total leukocyte count.

high male-to-female ratio observed in our study has also been seen in children, delays due to social factors may be less common on account
previous studies in other childhood tumors from our subcontinent.23 of parental care and attention. The prominent role of biologic drivers in
This highlights the presence of a gender gap in cancer diagnosis in India this age group may, thus, allow for a stronger association between short
that is not explained by tumor biology alone.2,23,24 symptom duration and adverse survival outcomes. However, while the
In our cohort, baseline metastatic disease and shorter duration of difference between symptom duration in the two cohorts was statis-
symptoms (<3 months) were identified as independent predictors of tically significant, the clinical relevance is questionable, as seen from
inferior OS. Ioris et al. also identified metastatic disease as a significant prior studies.28,29
prognostic factor.5 Metastases at presentation is known to be a pow- We found that the younger cohort presented with a favorable
erful predictor of survival, with long-term survival rates dipping down clinical profile in the form of less number of pelvic primaries, and a
from 70% in localized ES to as low as 9%–41% in metastatic ES.2,4,7,10,25 higher proportion of patients presenting with localized disease, both
The prognostic value of symptom duration prior to presentation of which have been associated with a better prognosis in ES.5,20
is controversial, likely due to interaction between social and bio- We however did not identify any difference in survival between the
logic drivers of delayed presentation.26 In non-Hodgkin’s lymphomas, younger and older cohorts in the metastatic and nonmetastatic sub-
quicker presentation to medical care following symptom onset has groups, signifying that age might not be independently predicting
been described as an adverse prognostic factor on account of its asso- outcomes. A further propensity-matched analysis confirmed that there
ciation with more aggressive tumor biology.27 However, in a cohort of was no significant difference between the younger and older cohorts.
bone sarcomas across age groups at our center, time to diagnosis was Therefore, we conclude that age is not an independent prognostic
not seen to have any impact on survival, likely denoting that tumor factor, although the results may be limited by our small sample size.
biology plays the dominant role in determining treatment outcomes, Prior to this study, multiple other groups have also shown mixed
rather than the symptom duration prior to presentation.26 In young results. The Polish sarcoma group,30 European Ewing Tumor Work-
15455017, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/pbc.31268 by Hacettepe Universitesi, Wiley Online Library on [16/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
NAGPAL ET AL . 7 of 9

TA B L E 3 Univariable and multivariable analysis of factors predictive of event-free survival in patients of age ≤5 years.

Univariable analysis Multivariable analysis

Factor Category Median (IQR) EFS HR 95% CI p-Value HR 95% CI p-Value


Sex Female (n = 26) 23.9 (9.2–NR) 1.11 0.57–2.19 .75
Male (n = 60) 26.8 (17.4–NR)
Fever at onset Yes (n = 28) 22.1 (10.4–NR) 1.14 0.57–2.25 .71
No (n = 56) 26.8 (15.8–NR)
Tumor diameter (cm) ≤5 (n = 21) NR (15.8–NR)
>5 (n = 49) 25.6 (16.4–NR) 1.39 0.63–3.09 .42
Site of disease Head and neck (n = 19) 45.7 (9.2–NR) 1
Extremity (n = 32) 19.7 (10.4–NR) 1.35 0.57–3.19 .68
Thorax (n = 24) NR (13.1–NR) 0.78 0.08–5.25 .61
Abdomen (n = 3) 22.3 (22.3–NR) 0.66 0.08–5.25 .69
Pelvis (n = 4) 16.4 (10.2–NR) 2.22 0.57–8.59 .25
Spine (n = 4) 13 (6.1–NR) 2.27 0.68–7.60 .68
Metastasis at baseline Yes (n = 25) 16.4 (8.4–26.8 months) 2.24 1.18–4.27 .014 2.47 1.29–4.76 .007
No (n = 59) 47.1 (21.5–NR)
Hemoglobin (g/dL) <10 (n = 52) 37.9 (13.6–NR) 1
≥10 (n = 28) 21.5 (10.4–NR) 1.29 0.66–2.53 .45
3
TLC (×10 /mm ) 3
>11 (n = 29) 22.1 (12.5-45.7) 1.46 0.77–2.78 .25
≤11 (n = 51) 26.8 (13.3–NR)
Symptom duration (months) ≥3 (n = 46) 26.8 (19.7–NR)
<3 (n = 37) 15.8 (12.5–47.1) 1.67 0.89–3.12 .10 1.89 0.99–3.57 .053
Tumor origin Skeletal (n = 63) 21.5 (13.6–NR)
Extra-skeletal (n = 23) NR (13.3–NR) 0.59 0.27–1.29 .19

Note: Column 3: median (IQR) or n (%).


Abbreviations: CI, confidence interval; cm, centimeters; EFS, event-free survival; g/dL, grams per deciliter; HR, hazard ratio; IQR, interquartile range; mm3 ,
per cubic meter; NR, not reached; TLC, total leukocyte count.

F I G U R E 2 (A) Kaplan–Meier curves for overall survival in patients without metastatic disease aged ≤5 versus >5 years. (B) Kaplan–Meier
curves for event-free survival in patients without metastatic disease aged ≤5 versus >5 years.
15455017, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/pbc.31268 by Hacettepe Universitesi, Wiley Online Library on [16/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
8 of 9 NAGPAL ET AL .

F I G U R E 3 (A) Kaplan–Meier curves for overall survival in patients with metastatic disease aged ≤5 versus >5 years. (B) Kaplan–Meier curves
for event-free survival in patients with metastatic disease aged ≤5 versus >5 years.

ing Initiative of National Groups 99 (EURO-EWING 99),10 EICESS,3 tent. Sandeep Agarwala; Vishesh Jain; Anjan Dhua; Devender Yadav;
and Children’s Oncology Group (COG)9 have concluded that young Bivas Biswas; Shah Alam Khan; Adarsh Barwad; Asit Ranjan Mirdha;
age is an independent predictor of a better EFS, whereas EW88,7 Ahitagni Biswas; and Deepam Pushpam conceptualized and designed
Wong et al.4 and Rodríguez-Galindo et al.8 have failed to conclude the study; and reviewed it critically for important intellectual con-
the same. tent; and edited the manuscript. Sameer Bakhshi conceptualized and
Our study is a large cohort study that provides data on outcomes designed the study; drafted the manuscript; reviewed it critically for
in children with ES belonging to the age group of ≤5 years. The use important intellectual content; edited the manuscript; and provided
of a uniform protocol allowed for better delineation of disease behav- administrative support. All authors have reviewed and approved the
ior in the absence of confounding factors. However, our study had final version of the manuscript to be published and agree to be account-
a few limitations. Our study did not assess differences in histomor- able for all aspects of the work in ensuring that questions related to
phology and immunohistochemistry in ES such as the presence of the accuracy or integrity of any part of the work are appropriately
Bcl-6 co-repressor (BCOR)-associated, CIC (Capicua)-rearranged and investigated and resolved.
EWSR1:NFATC2 (Ewings sarcoma RNA binding protein 1–nuclear fac-
tor of activated T cells 2) sarcomas in different age groups, which may ACKNOWLEDGMENTS
promote our understanding of the biologic basis of differences in ES The authors acknowledge every member of pediatric oncology team of
across age groups.31 The incorporation of social factors such as familial our center including research staff, nurses, and dietician for their exem-
income and maternal education may have allowed us to better grasp plary clinical services. In particular, the contributions of Ms Mamta
the relevance of biologic factors in the context of prevailing social Kumari who maintains and updates the patient database.
characteristics.
We found that young children with ES have distinct clinical CONFLICT OF INTEREST STATEMENT
presentation with less aggressive presenting characteristics. The authors declare no conflicts of interest.
However, younger age was not associated with better OS or EFS
after matching for all potential confounding baseline characteristics. DATA AVAILABILITY STATEMENT
A separate analysis of younger children with ES from prospective Anonymized patient-level data that support the findings of the study
trials may further enhance our understanding of the disease in this will be available from the corresponding author upon reasonable
age group. Future studies that look at gene expression profiling in request.
younger ES may help us to examine age-wise differences in disease
behavior at a molecular level, and identify potential therapeutic ORCID
targets that may allow us to further improve survival outcomes in Chitrakshi Nagpal https://orcid.org/0000-0002-7352-2446
this cohort. Shuvadeep Ganguly https://orcid.org/0000-0002-7296-6088
Deepam Pushpam https://orcid.org/0000-0001-8222-331X
AUTHOR CONTRIBUTIONS Sandeep Agarwala https://orcid.org/0000-0001-8315-5814
Chitrakshi Nagpal conceptualized the study; acquired; interpreted Vishesh Jain https://orcid.org/0000-0002-9273-097X
and analyzed the data; and drafted the manuscript. Archana Sasi; Anjan Dhua https://orcid.org/0000-0003-2050-7872
Shuvadeep Ganguly; and Vivek Kumar interpreted results; drafted Devender Kumar Yadav https://orcid.org/0000-0003-3158-1322
the work; and reviewed it critically for important intellectual con- Ahitagni Biswas https://orcid.org/0000-0002-3453-5340
15455017, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/pbc.31268 by Hacettepe Universitesi, Wiley Online Library on [16/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
NAGPAL ET AL . 9 of 9

Sameer Bakhshi https://orcid.org/0000-0001-9367-4407 17. Therneau TM. A package for survival analysis in R. R package ver-
sion 3.7-0. R-Project; 2023. Accessed October 14, 2023. https://cran.
r-project.org/web/packages/survival/index.html
REFERENCES
18. Kassambara A, Kosinski M, Biecek P, Fabian S. survminer: draw-
1. Cancer incidence and survival among children and adolescents— ing survival curves using “ggplot2.” R-Project; March 9, 2021.
pediatric monograph—SEER publications 1975–1995. SEER. Accessed Accessed October 14, 2023. https://cran.r-project.org/web/packages/
November 26, 2023. https://seer.cancer.gov/archive/publications/ survminer/index.html
childhood/index.html 19. Ho D, Imai K, King G, Stuart EA. MatchIt: nonparametric preprocess-
2. Sasi A, Ganguly S, Biswas B, et al. Development and validation of a ing for parametric causal inference. J Stat Softw. 2011;42:1-28. doi:10.
prognostic score at baseline diagnosis for Ewing sarcoma family of 18637/jss.v042.i08
tumors: a retrospective single institution analysis of 860 patients. Am 20. Biswas B, Agarwala S, Rastogi S, et al. High burden of metastases and
J Transl Res. 2022;14(2):927-941. poor outcome in pelvic PNET. Pediatr Blood Cancer. 2013;60(9):E97-
3. Cotterill SJ, Ahrens S, Paulussen M, et al. Prognostic factors in E99. doi:10.1002/pbc.24552
Ewing’s tumor of bone: analysis of 975 patients from the European 21. Biswas B, Thakar A, Mohanti BK, Vishnubhatla S, Bakhshi S. Prog-
Intergroup Cooperative Ewing’s Sarcoma Study Group. J Clin Oncol. nostic factors in head and neck Ewing sarcoma family of tumors.
2000;18(17):3108-3114. doi:10.1200/JCO.2000.18.17.3108 Laryngoscope. 2015;125(3):E112-E117. doi:10.1002/lary.24985
4. Wong T, Goldsby RE, Wustrack R, Cash T, Isakoff MS, DuBois SG. Clin- 22. Bakhshi S, Pathania S, Mohanti BK, Thulkar S, Thakar A. Therapy and
ical features and outcomes of infants with Ewing sarcoma under 12 outcome of primitive neuroectodermal tumor of the jaw. Pediatr Blood
months of age. Pediatr Blood Cancer. 2015;62(11):1947-1951. doi:10. Cancer. 2011;56(3):477-481. doi:10.1002/pbc.22615
1002/pbc.25635 23. Bhatia KP, Ganguly S, Sasi A, et al. Sex disparity in childhood cancer
5. Ioris MAD, Prete A, Cozza R, et al. Ewing sarcoma of the bone in chil- in India: a multi-centre, individual patient data analysis. Lancet Oncol.
dren under 6 years of age. PLoS One. 2013;8(1):e53223. doi:10.1371/ 2023;24(1):54-63. doi:10.1016/S1470-2045(22)00688-X
journal.pone.0053223 24. Bakhshi S, Biswas B, Mohanti BK, et al. Outcome and prognostic fac-
6. Worch J, Ranft A, DuBois SG, Paulussen M, Juergens H, Dirksen U. tors in metastatic primitive neuroectodermal tumors with uniform
Age dependency of primary tumor sites and metastases in patients chemotherapy protocol: a single center experience of 150 patients.
with Ewing sarcoma. Pediatr Blood Cancer. 2018;65(9):e27251. doi:10. J Clin Oncol. 2013;31(15):e21522-e21522. doi:10.1200/jco.2013.31.
1002/pbc.27251 15_suppl.e21522
7. Oberlin O, Deley MCL, Bui BN, et al. Prognostic factors in localized 25. Biswas B, Rastogi S, Khan SA, et al. Outcomes and prognostic fac-
Ewing’s tumours and peripheral neuroectodermal tumours: the third tors for Ewing-family tumors of the extremities. J Bone Joint Surg Am.
study of the French Society of Paediatric Oncology (EW88 study). Br J 2014;96:841-849. doi:10.2106/JBJS.M.00411
Cancer. 2001;85(11):1646-1654. doi:10.1054/bjoc.2001.2150 26. Sasi A, Ganguly S, Biswas B, et al. Determinants and impact of diag-
8. Rodríguez-Galindo C, Liu T, Krasin MJ, et al. Analysis of prognostic fac- nostic interval in bone sarcomas: a retrospective cohort study. Pediatr
tors in ewing sarcoma family of tumors. Cancer. 2007;110(2):375-384. Blood Cancer. 2023;70(3):e30135. doi:10.1002/pbc.30135
doi:10.1002/cncr.22821 27. Patel A, Tiwari A, Biswas B, Chand Sharma M, Vishnubhatla S, Bakhshi
9. Marina N, Granowetter L, Grier HE, et al. Age, tumor characteris- S. Clinical predictors and prognostic model for pediatric lymphoblas-
tics, and treatment regimen as event predictors in Ewing: a Children’s tic lymphoma treated with uniform BFM90 protocol: a single-center
Oncology Group report. Sarcoma. 2015;2015:1. doi:10.1155/2015/ experience of 65 patients from Asia. Clin Lymphoma Myeloma Leuk.
927123 2019;19(6):e291-e298. doi:10.1016/j.clml.2019.01.008
10. Ladenstein R, Pötschger U, Le Deley MC, et al. Primary disseminated 28. Nataraj V, Batra A, Rastogi S, et al. Developing a prognostic
multifocal Ewing sarcoma: results of the Euro-EWING 99 trial. J Clin model for patients with localized osteosarcoma treated with uniform
Oncol. 2010;28(20):3284-3291. doi:10.1200/JCO.2009.22.9864 chemotherapy protocol without high dose methotrexate: a single-
11. Biswas B, Rastogi S, Khan SA, et al. Hypoalbuminaemia is an inde- center experience of 237 patients. J Surg Oncol. 2015;112(6):662-668.
pendent predictor of poor outcome in metastatic Ewing’s sarcoma doi:10.1002/jso.24045
family of tumours: a single institutional experience of 150 cases 29. Biswas B, Rastogi S, Khan SA, et al. Developing a prognostic model for
treated with uniform chemotherapy protocol. Clin Oncol (R Coll Radiol). localized Ewing sarcoma family of tumors: a single institutional experi-
2014;26(11):722-729. doi:10.1016/j.clon.2014.05.006 ence of 224 cases treated with uniform chemotherapy protocol. J Surg
12. Jakutis G, Ragelienė L, Rascon J. Survival of children treated for Oncol. 2015;111(6):683-689. doi:10.1002/jso.23861
Ewing sarcoma in Lithuania: a single centre experience. Acta Med Litu. 30. Jagodzińska-Mucha P, Raciborska A, Koseła-Paterczyk H, et al. Age as
2017;24(4):199-208. doi:10.6001/actamedica.v24i4.3615 a prognostic factor in patients with Ewing sarcoma—the Polish Sar-
13. Karski EE, Matthay KK, Neuhaus JM, Goldsby RE, DuBois SG. Charac- coma Group experience. J Clin Med. 2021;10(16):3627. doi:10.3390/
teristics and outcomes of patients with Ewing sarcoma over 40 years jcm10163627
of age at diagnosis. Cancer Epidemiol. 2013;37(1):29-33. doi:10.1016/j. 31. Yoshida A. Ewing and Ewing-like sarcomas: a morphological guide
canep.2012.08.006 through genetically-defined entities. Pathol Int. 2023;73(1):12-26.
14. Martin RCG II, Brennan MF. Adult soft tissue Ewing sarcoma or doi:10.1111/pin.13293
primitive neuroectodermal tumors: predictors of survival? Arch Surg.
2003;138(3):281-285. doi:10.1001/archsurg.138.3.281
15. Womer RB, West DC, Krailo MD, et al. Randomized controlled trial
of interval-compressed chemotherapy for the treatment of localized
How to cite this article: Nagpal C, Ganguly S, Sasi A, et al.
Ewing sarcoma: a report from the Children’s Oncology Group. J Clin
Oncol. 2012;30(33):4148-4154. doi:10.1200/JCO.2011.41.5703 Ewing sarcoma among children 5 years of age or younger: Is it a
16. Grier HE, Krailo MD, Tarbell NJ, et al. Addition of ifosfamide and etopo- different disease? Pediatr Blood Cancer. 2024;e31268.
side to standard chemotherapy for Ewing’s sarcoma and primitive https://doi.org/10.1002/pbc.31268
neuroectodermal tumor of bone. N Engl J Med. 2003;348(8):694-701.
doi:10.1056/NEJMoa020890

You might also like