Rare Childhood cancer

Treating Rare Cancer in Children:
The Importance of Evidence
By Alberto S. Pappo, MD
Overview: The study of pediatric rare cancers, which account
for approximately 9% of all childhood malignancies, has been
hindered by their histologic heterogeneity and by their preferential
occurrence in adolescents, a population that has been
underrepresented in clinical trials sponsored by the National
Cancer Institute.
The use of cooperative group and investigator-initiated
registries can help improve our ability to identify and select
populations of patients with rare cancers that can benefit from
single-arm studies, and incorporation of biologic aims and
tissue banking can accelerate our understanding of the biology
of these cancers. These studies should be promoted
further through expansion of international outreach efforts.
Well-designed preclinical models that accurately recapitulate
human disease offer an attractive alternative to the study
of rare cancers and may accelerate the process of target
identification and drug discovery and development.
The concept of specialized clinics for selected rare cancers
has proven to be very successful in pediatric gastrointestinal
tumors. This paradigm should be further explored in other rare
cancers because it offers an unprecedented opportunity to
collaborate closely with interested investigators. In addition,
it offers patients an opportunity to discuss their disease with
specialists, allows these patients to provide tissue for further
research, and ultimately can promote the development of
clinical trials that are unique for that specific disease.
THE STUDY of rare pediatric cancers is complicated by
the fact that pediatric cancer itself is a rare occurrence,
with only approximately 13,000 patients younger than age
20 diagnosed yearly in the United States.1 Several groups
have tried to define a pediatric rare cancer; for example the
European Italian Tumori Rari in Eta Pediatrica (TREP)
(“the rare tumor project for pediatric patients”) project uses
a definition of a cancer that has an annual incidence of up to
2 per million and for whom there is no available clinical
trial.2 However, on the basis of data from the Surveillance,
Epidemiology, and End Results (SEER) program of the
National Cancer Institute, this definition would exclude
tumors such as thyroid cancer and melanoma, which have
an increasing incidence in adolescents and young adults, but
are still considered rare. With the merger of the Pediatric
Oncology Group and the Children’s Cancer Group in 2000,
the newly formed Children’s Oncology Group (COG) was
poised to increase our understanding of rare cancers in
children. With this objective in mind, a Rare Tumor Committee
was formed and the results of its initial experience
within the context of a cooperative group have been published
and summarized.3 When initially conceived, the main
objectives of the COG Rare Tumor Committee were to
facilitate the study of rare cancers, using the infrastructure
available at the time that the two Groups merged. To better
estimate the actual numbers of rare tumors diagnosed in the
United States, we opted to define them as those tumors
(mostly carcinomas) mostly diagnosed in older patients.
These subset of tumors closely fit the description of “other
malignant epithelial neoplasms and melanomas” listed in
the International Classification of Childhood Cancer subgroup
XI of the SEER database. These histologies include
nasopharyngeal carcinoma, adrenocortical carcinoma, melanoma,
nonmelanoma skin cancers, and other unspecified
carcinomas. We then used the available COG registry to
capture cases of these rare cancers and compared the registration
rates with the expected number of cases as calculated
according to available through the SEER database. We
were surprised to see that only 7% of expected cases of
selected histologies such as melanoma, thyroid carcinoma,
adrenocortical carcinoma, and nasopharyngeal carcinoma
were registered within COG. We also noted that the registration
rates varied according to histologic subtype and age,
with very low registration rates in patients with melanoma
and thyroid carcinoma and higher registration rates in
patients with retinoblastoma, adrenocortical carcinoma, and
nasopharyngeal carcinoma. These differences might be explained
in part by referral patterns that are dependent on
the age of the patient. Younger patients with adrenocrtiocal
carcinoma and retinoblastoma require multidisciplinary
care and are likely to be referred to tertiary academic
institutions for treatment. In contrast, adolescents with
thyroid carcinoma or melanoma are treated often with
primary surgical resection by professionals outside of a
pediatric cancer program.
Our initial mandate also included increasing the number
of banked “rare tumor” biologic specimens available for
future research through a COG banking specimen protocol,
activated in October of 2003. The number of samples received
by this repository over a 4-year period totaled 517
snap-frozen specimens, of which only 56 (11%) were considered
to belong to a rare cancer. Finally, the COG Rare
Tumor Committee promoted the development of clinical
trials, and two COG-sponsored trials for the treatment of
nasopharyngeal carcinoma and adrenocortical carcinoma
were developed. The enrollment rates for both studies were
less than initially projected, and the statistical sections had
to be modified in order to ensure that these trials could be
completed in a timely manner.3 Thus our preliminary experience
with the study of rare cancers within the context of a
pediatric cooperative group was somewhat disappointing
and highlighted the difficulties associated with the study
of this very challenging group of pediatric tumors. Other
opportunities or avenues for research that could potentially
increase our knowledge in rare cancers, as well as methods
for evidence gathering and interpretation must be considered.
From the Division of Oncology, St. Jude Children’s Research Hospital, Memphis, TN.
Author’s disclosure of potential conflicts of interest are found at the end of this article.
Address reprint requests to Albert S. Pappo, MD, St. Jude Children’s Research Hospital,
262 Danny Thomas Place, Memphis, TN 38105; email: alberto.pappo@stjude.org.
© 2012 by American Society of Clinical Oncology.
1092-9118/10/1-10
586
Registries
The use of registries to study rare cancers has become
commonplace within the setting of national and international
cooperative groups.2-5 For example, TREP in Italy
was launched in 2000 and accrued more than 300 eligible
patients with a variety of histologies over a 6-year period.6
In their experience, thyroid carcinoma, carcinoma of the
appendix, and gonadal germ cell tumors were the most
common tumors registered, and significant underreporting
of various rare cancers was also found among adolescents.6
The TREP project has published valuable information regarding
the natural history, treatment, and outcome of
selected rare cancers and has conducted a prospective trial
for the treatment of nasopharyngeal carcinoma.7 In an effort
to expand the registration efforts within the Children’s
Oncology Group, the Children’s Cancer Research Network
(CCRN) was launched in 2007. The purpose of this initiative
was to integrate the registration process for patients with
tumors at COG member institutions into the research of the
group, providing a research resource in North America. The
main purpose of this trial was to obtain informed consent
from patients—infant and children, adolescent, and young
adults with newly diagnosed cancers—and securely record
patient identification and contact information. In addition,
data concerning their cancer and consent for permission for
future contact with the patients for possible epidemiologic
studies is also obtained. In contrast with the CCRN, SEER
does not ascertain all cases of childhood cancer and does not
allow for future contact of research participants, limiting the
availability of epidemiologic studies in these patients. A
preliminary analysis of enrollments of rare cancers to the
CCRN study demonstrates that the registration rates for
adrenocortical carcinoma, thyroid carcinoma, nasopharyngeal
carcinoma, and melanoma have not significantly
changed, with approximately two-thirds of the estimated
incidence cases of patients with adrenocortical carcinoma
being enrolled but only approximately 5% of patients with
melanoma being enrolled on the registry. In an attempt to
more efficiently define the natural history and biology or
rare cancers, various groups have developed independent
registries that have provided meaningful insights into the
biology of pediatric rare cancers. The International Pediatric
Adrenocortical Tumor Registry was founded in 1990 and has
described the clinical characteristics and outcome of 254 of
children with this rare disease.8 Their findings provided the
basis for the development of a collaborative cooperative
group trial between COG and Brazilian institutions. This
trial assesses the efficacy of surgery in stage I disease, the
role of retroperitoneal lymph node dissection in stage II
disease, and the role of multimodal chemotherapy with
mitotane and cisplatin-based regimens in advanced-stage
disease. In addition, the trial will continue to expand on the
initial observations of the registry that TP53 germ-line
mutations appear to occur in up to 70% of cases (one-third
being de novo) and that a unique TP53 Arg337 mutation
affects children in Southern Brazil. The latter does not show
the hereditary pattern seen in Li–Fraumeni syndrome and
has different transactivation activity that may disrupt the
function of p53 in a pH dependent manner.9,10 In another
independent effort, investigators of the pleuropulmonary
blastoma registry have identified heterozygous germ-line
mutations of DICER1, an endoribonuclease that is essential
for processing micro RNAs in 10 families with familial
pleuropulmonary blastoma.11 More recently, the mutational
spectrum of this gene has been expanded to include other
tumors and associations including ovarian Sertoli-Leydig
cell tumors, cystic nephroma, thyroid cysts, multinodular
goiter, and embryonal rhabdomyosarcoma.12-15
Preclinical Models
Well-designed comprehensive preclinical studies using
genetically engineered and orthotopic xenografts that
closely recapitulate the molecular, cellular, and genetic
features of human tumors can be an invaluable tool for
identifying novel targets and promising therapies for rare
cancers. In adults, the use of preclinical models successfully
identified everolimus and sunitinib as active agents for the
treatment of pancreatic neuroendocrine tumors.16,17 Retinoblastoma
is a rare cancer of the retina that begins in utero
and is usually diagnosed during the first few years of life.
There are approximately 300 cases in the United States each
year and approximately 5,000 diagnosed world-wide. Preclinical
studies have determined that the p53 gene is intact
but the pathway is silenced by overexpression of MDMX.18
Preclinical models have demonstrated that the subconjunctival
administration of nutlin-3a, a small molecule inhibitor
of the MDMX-p53 interaction in combination with systemic
topotecan is effective in killing and reducing the tumor
burden of retinoblastoma cells in culture, as well as in
genetic and human orthotopic xenograft models of retinoblastoma.
19 These studies provided direct evidence that
targeting the p53 pathway in vivo with MDMX inhibitors is
feasible and provides a potential novel therapy for this
disease. Similarly, using an integrated epigenetic and gene
expression analysis, SYK has been shown to be an important
oncogene in retinoblastoma. Furthermore, the use of SYK
inhibitors such as BAY-613606 and R406 cause cell death;
when combined with topotecan, this combination significantly
improves the outcome of mice.20 These observations
suggest that well-designed preclinical studies can identify
novel targets and prioritize the development of new agents
for rare cancers.
Prospective Clinical Trials
The small number of patients available for enrollment in
pediatric rare cancer trials prevents the design of prospec-
KEY POINTS
● Rare cancers in children are difficult to study.
● Registries and single-arm trials can help develop
standards of care and increase the numbers of biologic
specimens, which could further facilitate study
of these tumors.
● Preclinical models offer a unique mechanism for
rapidly identifying novel druggable targets and for
prioritizing new therapies.
● Specialized clinics can foster collaborations that will
improve the care and study of pediatric rare cancers.
● International collaborative efforts must be expanded
in order to speed up progress in the field of pediatric
rare cancers.
RARE CANCER IN CHILDREN
587