Keywords
acute myeloid leukemia - India - lower-middle-income countries - abandonment
Introduction
Acute leukemia accounts for approximately one-third of all childhood malignancies,
of which 15 to 20% cases comprise of acute myeloid leukemia (AML).[1] The outcomes of childhood AML in high-income countries (HICs) have currently surpassed
70% with an increased focus on targeted therapies to further these outcomes and also
simultaneously reduce toxicity.[2]
[3] Lower-middle-income countries (LMICs) continue to have suboptimal outcomes due to
various socioeconomic and disease-related factors.[3] There is limited data on childhood AML from India.[4] As a result, there is limited understanding of disease trends, which may ultimately
compromise patient care. A previous systematic review from India, which included studies
published between 1990 and 2010, highlighted several shortcomings of managing pediatric
AML.[5] The current systematic review was undertaken to study the treatment strategies and
outcomes of pediatric AML in India. The review included studies published between
January 1, 2011 and December 31, 2021.
Materials and Methods
Protocol and Registration
This systematic review was registered on PROSPERO (ID42021273218).
Eligibility Criteria
Inclusion Criteria
-
Studies reporting on pediatric AML in India.
-
Studies written in English.
-
Prospective, retrospective, and ambispective studies.
Exclusion Criteria
-
Studies on pediatric AML not from India.
-
Case reports, reviews, and books.
Settings
There were no restrictions on the type of setting in which the studies were conducted.
Information Source
A systematic search of the MEDLINE, Google Scholar, and SCOPUS database for published
studies on pediatric AML from India was conducted. In addition, SIOP conference abstracts
were also screened. The reference lists of the included studies or relevant reviews
were screened for other eligible studies.
Time
Search of database was from January 1, 2011 till January 31, 2021. SIOP conference
abstracts were screened from year 2011 to 2020.
Literature Search
A comprehensive literature search was performed using text words “Acute myeloid leukemia,”
“AML,” “child*,” “India.” Articles published in English alone were reviewed. Literature
search was as per preferred reporting items for systematic reviews and meta-analyses
(PRISMA) statement. The full search strategy is shown in [supplementary material].
Study Selection
Two review authors (S.S and V.R.M.G) independently screened the titles and abstracts
yielded by the search against the inclusion and exclusion criteria. Full reports for
all titles and abstracts were obtained if they appeared to meet the inclusion criteria
and in case of any uncertainty. Review authors then screened the full text reports
and decided whether the inclusion criteria were met. If necessary, additional information
from study authors was sought to resolve questions about eligibility and disagreement
was resolved through discussion. Reasons for excluding trials were also recorded.
None of the review authors were blinded to the journal titles or to the study.
Data Collection Process
Data extraction from the included studies was performed using standardized data collection
forms. Two reviewers (S.S and N.D) independently extracted the data to reduce the
bias and errors in data extraction and the studies in question were jointly reviewed
by the two investigators and the final determination was reached by consensus.
Data Items
The information that was extracted from each study included surname of the first author,
year of study, median/mean age with range, number of patients, chemotherapy administered,
induction mortality, complete remission (CR) rate, duration of follow-up, relapse,
event-free survival (EFS), overall survival (OS), treatment-related mortality (TRM),
treatment abandonment, prognostic factors, and use of hematopoietic stem cell transplant
(HSCT).
Evaluation of Quality and Risk of Bias
Quality was assessed by two authors using the quality assessment tool for observational
cohort and cross-sectional studies from the National Heart, Lung, and Blood Institute
(NHLBI) of the National Institutes of Health.[6]
Synthesis Method
All studies included were screened for the required data items and results were tabulated
using Microsoft Word software. Categorical variables were expressed as the number
of cases and percentages (%). Mean along with 95% confidence interval (CI) was calculated
to report the incidence of TRM and abandonment rates. Statistical analyses were done
using the R software version 4.0.2.
Results
Literature Search
A total of 1,057 studies and 15 SIOP conference abstracts were obtained after the
initial search. Additionally, seven other studies were added after citation searching.
After removing duplicates and screening the titles and abstracts of the publications,
full text of 60 studies were assessed of which 19 were included for the systematic
review. The PRISMA flowchart is shown in [Fig. 1].
Fig. 1 Flow diagram of the systematic review according to preferred reporting items for
systematic reviews and meta-analysis (PRISMA) guidelines.
Quality of Studies
The quality assessment tool for observational cohort and cross-sectional studies from
the National heart, Lung, and Blood Institute of the National Institutes of Health
was adapted to assess the quality of included studies ([Supplementary Table S1]).[7] Overall, the quality of the study was poor in 1 (5%) study, fair in 8 (42%) studies,
and good in 10 (53%) studies of the 19 studies included in the systematic review.
Characteristics of the Studies
A total of 1,210 patients were included from the 19 studies.[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26] Three of the 19 studies also included patients with acute promyelocytic leukemia
(APML).[19]
[21]
[22] Eight studies were published before 2015, while the remaining 11 studies were published
in or after 2015. The various studies included patients between 1 and 19 years of
age. There was slight predominance of males across majority of the studies. The salient
features of the studies are summarized in [Table 1].
Table 1
Characteristics of 19 studies included in the systematic review
|
Study
|
Type of study
|
Time period
|
Age (in years)
|
M:F ratio
|
Number of patients
|
Chemotherapy
|
Risk group included
|
Median follow-up
|
Abandonment (%)
|
CR (%)
|
Relapse/refractory disease (%)
|
EFS/DFS
|
Induction mortality (%)
|
Overall TRM (%)
|
OS
|
|
Gupta et al 2011
[8]
|
Retrospective
|
2005–2009
|
Mean: 12.4 (1–18)
|
1.9:1
|
35
|
In-I: 3 + 7
(D45mg/m2 + Arac 100mg/m2)
In-II: HAM
Consol: HIDAC
|
All
|
NA
|
5.7%
|
77.1%
|
40%
|
2 years DFS: 40%
|
2.9%
|
5.7%
|
NA
|
|
Yadav et al 2011
[9]
|
Retrospective
|
2005–2010
|
NA
|
NA
|
51
|
UKAML12 protocol
|
All
|
NA
|
55%
|
NA
|
26%
|
NA
|
22%
|
48%
|
26%
|
|
Mohammed et al 2013
[10]
|
Retrospective
|
2006–2013
|
NA
|
NA
|
34
|
NA
|
All
|
NA
|
26%
|
56%
|
26%
|
NA
|
12%
|
NA
|
59%
|
|
Kota et al 2013[11]
|
Retrospective
|
2007–2012
|
1–19
|
1:6:1
|
63
|
In: 3 + 7
Consol: NA
|
NA
|
11 months
|
21%
|
78%
|
NA
|
Median EFS: 11 months
|
NA
|
NA
|
3 year OS: 15%
|
|
Sharawat et al 2014
[12]
|
Retrospective
|
2008–2010
|
Median: 10(1–18)
|
3:1
|
64
|
In:
3 + 7
(60mg/m2 *3 days)
Consol: HIDAC
|
All
|
18.3 months
|
NA
|
83%
|
NA
|
EFS: 30.2 ± 5.8%
DFS:
43.03 ± 7.3%
|
NA
|
NA
|
37.1 ± 6.3%
|
|
Jain et al 2014
[13]
|
Retrospective
|
2000–2013
|
NA
|
NA
|
88
|
In: 3 + 7 & 5 + 2
Consol: HIDAC + M
|
NA
|
NA
|
34%
|
NA
|
NA
|
NA
|
NA
|
18%
|
NA
|
|
Jayabose et al 2014
[14]
|
Retrospective
|
2010–2014
|
NA
|
0.9:1
|
39
|
Modified MRC-10 protocol + M
|
All
|
29 months
|
NA
|
72%
|
21.4%
|
3 year EFS:
40%
|
NA
|
18%
|
3 year OS: 47.5%
|
|
Siddaiahgari et al 2014
[15]
|
Prospective + Retrospective
|
2009–2012
|
|
0.7:1
|
32
|
UK AML 15 protocol
|
All
|
NA
|
NA
|
94%
|
16%
|
NA
|
NA
|
6%
|
72%
|
|
Philip et al 2015[16]
|
Retrospective
|
2012–2014
|
NA
|
NA
|
23
|
AML-BFM 98 protocol + M
|
NA
|
7 months
|
NA
|
NA
|
NA
|
NA
|
17%
|
NA
|
1 year OS:
70.4 ± 10.7%
|
|
Radhakrishnan et al 2015
[17]
|
Retrospective
|
2008–2013
|
Median: 9 (1–17)
|
2.25:1
|
72
|
In: DAE/DA
Con: HIDAC
|
All
|
11.7 months
|
NA
|
72%
|
NA
|
EFS: 28%
|
5.5%
|
7%
|
OS: 36%
|
|
Ramamoorthy et al 2015
[18]
|
Retrospective
|
2004–2013
|
Mean: 7.3 ± 3.6
|
3.2:1
|
100
|
AML MRC 12 protocol
|
All
|
NA
|
3%
|
64%
|
25%
|
DFS: 34.7%
|
25%
|
48%
|
27.2%
|
|
Seth et al 2016
[19]
|
Retrospective
|
2011–2015
|
Median: 7.5 (1.5–13)
|
NA
|
71
(Included APML)
|
MRC-10 protocol
|
All
|
NA
|
25%
|
95%
(excluding APML)
|
NA
|
3 year EFS: 43%
(excluding APML)
|
5.4%
|
27%
|
3 year OS: 55%
(excluding APML)
|
|
Narula et al 2017
[20]
|
Retrospective
|
2011
|
NA
|
NA
|
65
|
In: 3 + 7
Consol: HIDAC
+ M
|
All
|
NA
|
NA
|
NA
|
NA
|
3 year DFS: 66%
3 year EFS: 49%
|
<20%
|
NA
|
3 year OS:66%
|
|
Naseer et al 2017
[21]
|
Retrospective
|
2012–2017
|
NA
|
2:1
|
42
(included APML)
|
In: 7 + 3 and 5 + 2
Con:
HIDAC
+ M
|
All
|
NA
|
9.4%
|
56%
|
25–28%
|
NA
|
18%
|
NA
|
19%
|
|
Kapoor et al 2018[22]
|
Retrospective
|
2015–2018
|
NA
|
NA
|
24
(Included APML)
|
In: 3 + 7
Consol: HIDAC
|
All
|
31 months
|
4%
|
NA
|
29%
|
NA
|
NA
|
NA
|
67%
|
|
Peyam et al 2018
[23]
|
Retrospective
|
2011–2017
|
Mean: 6.96(1–12)
|
2.2:1
|
114
|
MRC 15 protocol
|
All
|
NA
|
8.8%
|
67.5%
|
22.8%
|
3 year EFS: 31.6%
|
NA
|
30.7%
|
NA
|
|
Sinha et al 2019[24]
|
Retrospective
|
2014–2015
|
<15
|
1.7:1
|
65
|
NA
|
NA
|
NA
|
20%
|
NA
|
NA
|
NA
|
NA
|
NA
|
36.9% at 5 months after diagnosis
|
|
Uppuluri et al 2020[25]
|
Retrospective
|
2002–2019
|
8
|
NA
|
48
|
MRC 15 protocol
|
All
|
NA
|
NA
|
NA
|
41%
|
NA
|
6.2%
|
NA
|
5 year OS:53%
|
|
Srinivasan et al 2020
[26]
|
Retrospective
|
2014–2017
|
9
|
NA
|
180
|
Upfront OMCT f/b 3 + 7 and HIDAC + M
|
All
|
25 months
|
NA
|
NA
|
NA
|
2 year EFS:
46–52
|
6.5%
|
NA
|
2 year OS:
47–53%
|
Abbreviations: APML, acute promyelocytic leukemia; AraC, cytarabine; consol, consolidation;
CR, complete remission; DAE, Daunorubicin, Cytarabine, Etoposide; Dauno, daunorubicin;
DFS, disease-free survival; EFS, event-free survival; f/b, followed-by; HAM, high-dose
cytarabine, mitoxantrone; HIDAC, high-dose cytarabine; In, induction; M, maintenance;
M:F, male:female; NA, not available; OMCT, oral metronomic chemotherapy; OS, overall
survival; TRM, treatment related mortality.
Data regarding induction chemotherapy were available from 17 studies.[8]
[9]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[25]
[26] Anthracycline-based regimens were used for induction in all studies. 3 + 7 regimen
consisting of daunorubicin (45–60 mg/m2/day) and cytarabine (100–200 mg/m2/day) was the standard regimen used in nine studies, while MRC AML based regimens
that included an additional third agent (etoposide) were used in seven studies. In
another study, patients were treated with AML-Berlin-Frankfurt-Munich (BFM) 98 protocol
in which idarubicin was the anthracycline used. CR rate was available from 11 studies
and varied between 56 and 95%, and overall there was no major difference in CR rates
between 3 + 7 regimens (56–78%) and MRC-based regimens (64–95%)[7]
[9]
[10]
[11]
[13]
[14]
[16]
[17]
[18]
[20]
[22]. Six of the 17 studies used maintenance chemotherapy.[13]
[14]
[16]
[20]
[21]
[26] Six studies mentioned regarding HSCT (Reference: 8,9,16,17,22,25). A total of 20
patients underwent HSCT in these studies, 15 in CR1 and another 5 in CR2.[8]
[9]
[16]
[17]
[22]
[25] Overall induction mortality (10 studies) and TRM (9 studies) were 12% (95% CI: 6.4–17.8)
and 23.2% (95% CI: 10.3–35.9), respectively.[7]
[8]
[9]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[22]
[24]
[25] Only three studies had a TRM of less than 10%, while four other studies had a high
TRM over 20%.
Duration of follow-up was available from seven studies and the shortest and longest
follow-up period was 7 and 31 months, respectively.[11]
[12]
[14]
[16]
[22]
[26]
[27] Data pertaining to event-free survival/disease-free survival (EFS/DFS) was available
from 10 studies.[8]
[11]
[12]
[14]
[17]
[18]
[19]
[20]
[23]
[26] The overall EFS/DFS reported among these studies ranged between 28 and 52%. Data
for OS was available from 16 studies.[8]
[9]
[10]
[11]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[23]
[24]
[25] Philip et al reported an OS of 70% with a follow-up of 7 months. Remaining 15 studies
had OS ranging between 15 and 66%. In general, the OS of studies that used maintenance
therapy (19–66%) was not different from those studies that did not offer maintenance
(15–55%). Prognostic factors could be determined from four studies.[8]
[12]
[17]
[21] High-risk cytogenetics that included -5/del 5q, -7/del 7q, complex cytogenetics
(defined as more than 3 structural and/or numerical abnormalities) were cited to have
negative impact on CR and relapse rate. Sharawat et al highlighted the negative impact
FMS-like tyrosine kinase 3 - internal tandem duplication (FLT3-ITD) mutations (DFS
of 18% for FLT3-ITD-positive vs. 51% for FLT3-ITD-negative patients).[12] Kapoor and Yadav in their paper highlighted that the negative impact of adverse
cytogenetic/molecular can be negated by HSCT in CR1.[22] The mean incidence of treatment abandonment that was available from 11 studies was
19.3% (95% CI: 10.9–27.5).[7]
[8]
[9]
[10]
[12]
[17]
[18]
[20]
[21]
[22]
[23] Six (55%) of these studies reported an abandonment rate over 20%.
Discussion
Treatment of AML in children continues to remain a challenge in LMICs. A previous
systematic review published by Kulkarni and Marwaha in the year 2010 summarized two
decades of experience of treating pediatric AML in India.[5] Their review included 322 children between the year 1990 and 2010, which is much
smaller than our current review that included 1,200 children treated over a shorter
duration of 10 years. Also, a recent systematic review on pediatric AML from LMICs
acknowledged that maximum data was contributed from India.[3] This is a step in the right direction indicating that more Indian children with
AML are being treated and reported. But, the true incidence of childhood AML in India
is unknown. According to World Health Organization, the estimated number of new cases
of leukemia from India, in the 0 to 14 age group, for the year 2020 was 11,850 and
considering that approximately 15% of these patients have AML, the annual incidence
of childhood AML should be approximately 1,750.[28] Thus, there continues to be underreporting and underdiagnosis of pediatric AML in
India.
Treatment abandonment is a major hurdle and is one of the most common reasons for
treatment failure in LMICs.[29] In fact, treatment abandonment is thought to contribute to at least a third of the
survival difference between HICs and LMICs.[30] Though not systematically reported, the current review highlights an alarmingly
high abandonment rates among children with AML, with no major improvements in comparison
to previous reports.[5] Perceived prognosis of the disease, cost of treatment, and concerns of toxicity
are few of the contributing factors to such high abandonment rates. Studies from India
and other LMICs have highlighted that a comprehensive support group consisting of
clinicians, as well as existing non-governmental organizations and governmental organizations
can significantly reduce abandonment.[31]
[32]
[33]
TRM is the next biggest hurdle in the treatment of pediatric AML in LMICs. The standard
of care for pediatric AML continues to be anthracycline-based induction followed by
three to four cycles of consolidation. Most of the chemotherapy protocols for treating
pediatric AML in India have been adopted from HICs, but the lack of essential supportive
care and option of intensive care unit admission, which are considered to be indispensable
during intensive AML treatment, have led to survival gap in comparison to HICs. For
example, Yadav et al highlighted a very high TRM of 48% when treated with the UKAML12
protocol.[9] On the contrary, the original UKAML12 trial that used the same protocol had a TRM
of only 10%.[34] Similar to previous studies from India, the current review estimated a high incidence
of induction mortality (12%) and overall TRM (23%), which is much higher compared
with 5 to 10% occurring in HICs.[4]
[5]
[35]
[36]
[37] High rates of infection with multidrug-resistant organisms, invasive fungal infections,
and poor nutritional status have led to poor tolerance and subsequently a high TRM
during intensive chemotherapy. Uppuluri et al highlighted that early intervention
by the pediatric intensive care team and granulocyte transfusion positively impacts
survival.[25] For patients in resource-limited settings with level two facilities, SIOP Pediatric
Oncology in Developing Countries (PODC) guidelines recommend an alternative strategy,
to begin treatment with inexpensive, low-intensity oral chemotherapy followed by low-dose
or standard-dose induction to reduce TRM and abandonment.[38] For patients with baseline adverse host-related factors, use of upfront low-dose
oral chemotherapy as a bridge to intensive chemotherapy has been shown to be safe
and reduce TRMs with comparable outcomes to those who directly receive intensive chemotherapy.[26] Another modifiable factor that contributes to TRM is malnutrition. The reported
incidence of malnutrition among Indian children with leukemia is approximately 50%.[27]
[39]
[40] Promoting routine nutritional assessment and ensuring availability of nutritional
supplements that are affordable and culturally appropriate, such as ready-to-use therapeutic
foods, must be incorporated into the care of childhood leukemia in India.
Lack of uniform access and high cost contribute to low rates of HSCT in India. In
the current review, HSCT rate was less than 2%. This remains a significant concern
for a disease like pediatric AML in which a third of the patients are thought to be
high risk and thus would qualify for HSCT in CR1. Also, patients from LMICs do not
have access to many of the newer therapies such as antibody drug conjugates and small
molecule inhibitors, which are currently available for AML. Incorporating a simple
risk stratification, which will otherwise identify the favorable risk group who can
be cured by chemotherapy alone, will be helpful, especially in the setting of limited
access to HSCT. Identifying certain high-risk mutations such as FLT3 can be of therapeutic
benefit in light of access to targeted therapies such as tyrosine kinase inhibitors.
The survival of pediatric AML in HICs has reached 70% and is mostly attributed to
advanced diagnostic techniques, better supportive care, and improved salvage options
including HSCT. This has not been the scenario in India and other LMICs where the
OS ranges between 10 and 50%.[3] Compared with previous studies published in India, our current review shows no major
improvement in survival trends in the past 30 years.[5] The lower survival rate in India can be attributed to high TRMs, high treatment
abandonments, and low salvage rate after relapse. While the HICs continue to improve
upon the benchmark survival of 70% through refinement of molecular risk stratification
and increased efforts toward personalized targeted therapy approaches, the immediate
steps in LMICs must address both socioeconomic and disease-related challenges as discussed.
The current systematic review has certain limitations. Of the 19 studies included,
9 studies were published only in abstract format and 3 studies included APML patients
that are often analyzed as a separate subset. Certain characteristics including baseline
comorbidities, risk stratification, and delay in diagnosis were not captured. The
median follow-up time was not mentioned in majority of the studies and among those
studies which mentioned it, three had a follow-up duration of less than 1 year.
Conclusion
In conclusion, the treatment outcomes of pediatric AML in India are substantially
inferior compared with HICs. Lowering TRM and abandonments is of utmost importance.
A holistic approach of including a social support team, intensified patient counselling,
ensuring uniform access to cancer therapy and supportive care will go a long way in
improving the outcomes of pediatric AML in India. Collaboration and prospective multicenter
studies may not only ensure standard of care treatment but also reduce abandonment
rates. The Indian Pediatric Oncology group (InPOG) initiative is a step in that direction.[41]