Discussion
Historically, all embryonal tumours originating in the CNS were grouped
under the umbrella term- PNET, irrespective of the site of tumour. They
had similar appearance under microscope, consisting of small round blue,
undifferentiated neuroepithelial cells, usually with high mitotic rate
and had neuroectodermal origin. Even some tumours outside CNS,
originating from the neural crest, were included and called pPNET, which
resembled histomorphologically with Ewing sarcoma (ES) of bones and
extra-osseous ES. Also, a reciprocal translocation t(11;22)(q24;q12) was
found in more than 95% cases of both pPNETs and ESs, using cytogenetic
analysis, so both were grouped together under Ewing sarcoma family of
tumours (ESFT)(3). Intracranial ES/pPNET usually arise from meninges,
while cPNETs have neuroparenchymal origin. While both cPNET and ES/pPNET
have aggressive courses, the former rarely metastasises outside the CNS
and the latter has higher metastatic potential, usually to bones and
lungs. Treatment protocols also differ between the two types of PNET;
CSI with focal boost RT and chemotherapy are needed for cPNET after its
surgical debulking and for ES/pPNET, post-operative chemotherapy on the
lines of ES followed by adjuvant local RT is sufficient. Hence,
differentiation between the two entities is of utmost importance,
especially in case of overlapping areas of CNS and peripheral nervous
system (PNS) like meninges and spinal canal. The survival rates, though,
are similar if treated as per the standard protocols, ranging from
50-70% for localized diseases(4).
The differentiation, not feasible by morphology, is done using CD99
(MIC2 glycoprotein) IHC and FISH for EWSR1 gene rearrangement. The
membranous expression of CD99 serves as a highly reliable and sensitive
diagnostic indicator for primary intracranial ES/pPNETs, positive in
almost all cases and negative in cPNET(5,6). It is not advised as a
specific IHC marker for diagnosing ES/pPNETs due to its positivity in
other small, blue round cell tumours like lymphoblastic lymphomas,
ependymomas, and rhabdomyosarcomas, even though the staining pattern in
these tumours often appears cytoplasmic rather than the characteristic
membranous staining seen in ES/pPNETs. The membrane protein FLI-1 is
typically present in ES/pPNETs and employing both CD99 and FLI-1 IHC
proves beneficial in the diagnosis(7). The gold standard to diagnose
ES/pPNET and rule out cPNET, however is molecular testing to depict
EWSR1 gene rearrangement. ES/pPNET has a characteristic translocation
between EWSR1 gene on chromosome 22 and one of the ETS family of genes,
most commonly FLI1 (chromosome 11), i.e., t(11;22)(q24;q12), and also
ETV1 on chromosome 7 and ERG on chromosome 21. FISH assays using EWS
break-apart probes are around 91-100% sensitive and specific; RT-PCR is
used particularly to identify the partner gene for EWSR1, with a 67%
concordance with FISH assays(8,9). The bottom line is that all
morphologically diagnosed intracranial (and spinal canal) ES/PNET must
be subjected to CD99 IHC and EWSR1 gene rearrangement using FISH to get
a clear picture of the diagnosis and accordingly, plan for the
treatment.
The initial theories considered CNS PNETs and MBs to be the same
disease, arising in different locations, former arising supratentorial
and the latter, infratentorial(10). Slowly, both were accepted to be
different biologically(11). The 2007 WHO classification of CNS tumors
included CNS-PNET not otherwise specified (NOS) and four morphologically
distinct CNS-PNET variants - Medulloepithelioma (ME), CNS
ganglioneuroblastoma, CNS neuroblastoma, and ependymoblastoma (EB). CNS
PNETs were then recognized as a molecularly heterogeneous group, with
the need for better classification. Based on the expression of cell
lineage markers, LIN28 and OLIG2, 3 molecular subgroups were identified
- primitive neural, oligo-neural and mesenchymal(1). DNA methylation
profiling identified four molecular entities under CNS-PNET-CNS
neuroblastoma with FOXR2 activation (CNS NB-FOXR2), CNS Ewing sarcoma
family tumor with CIC alteration (CNS EFT-CIC), CNS high-grade
neuroepithelial tumor with MN1 alteration (CNS HGNET-MN1), and CNS HGNET
with BCOR alteration (CNS HGNET-BCOR)(12). Top of Form
Bottom of Form
The term PNET has been scrapped by WHO ever since WHO classification of
CNS tumours 2016 due to the advent of molecular classification(13). As
per the latest WHO classification of CNS tumours 2021, these tumours are
under the classification of embryonal tumours, which are broadly divided
into MBs and other CNS embryonal tumours. The latter include a host of
tumours with different molecular and histological hallmarks, and a
subgroup called CNS embryonal tumour, which was devoid of any such
hallmark, i.e., not otherwise specified (NOS) and not elsewhere
classified (NEC)(14). So, the erstwhile intracranial central PNET
(cPNET) would now come under CNS embryonal tumour. cPNET account for
3-5% of paediatric CNS tumours and are at least five times less common
than MBs. cPNET occur mainly in the cerebral hemispheres, the most
common region being frontoparietal(15). Though biologically different,
pinealoblastomas (PB) are sometimes included under cPNET and account for
20% of cPNET(16). cPNETs are usually diagnosed in young children,
mainly less than 5 years old(17,18). Staging investigations are usually
the same as those for MBs, although the clinical significance and
correlation with long term disease control, especially that of the
extent of resection, is less clear than for MB(17,18). The survival is
even poorer compared to MB. For average-risk, the five-year
progression-free survival (PFS) is 50% and >80%
respectively and for high-risk, <20% and 50-60%
respectively(19). Despite differences in outcome, treatment is still
done on the lines of high-risk MBs in children as well as adult
patients, due to the rarity of these tumors(20,21).
Given the fact that up to 33-35% of cPNETs have CSF spread at
diagnosis, lumbar puncture and CSF cytology is a very crucial step in
staging and evaluation of CNS-PNETs(22,23). Another point worth
mentioning is the 3.3% chance of brain metastases and 9% chances of
skull bone metastases in ESs(24,25). Hence, PET-CT has some role in
ruling out occult extracranial primary ES with intracranial metastases.
F-18 FDG PET-CT has been shown to localise the primary disease in
carcinoma of unknown primary (CUP) in 40% cases, with the detection
rate being 77% in case of CUP with brain lesions(26,27).
The surgical approach has been maximum safe resection with the aim of
gross total resection (GTR). The improvement in outcomes in patients
with minimal residual disease has been well demonstrated in
non-disseminated MBs, though similar evidence is lacking for cPNETs(28).
It should also be the aim of surgery to maintain neurologic function.
Surgery is conventionally followed by radiotherapy and chemotherapy. RT
guidelines are similar to high-risk MB, though the need of CSI in
non-disseminated cPNET has never been proven and focal RT has been tried
in well-localised lesions post GTR(19). Chemotherapy is also planned on
similar lines as high-risk MB, with concurrent and adjuvant
chemotherapy. Multiple attempts have been done to further intensify the
treatment in view of poor outcomes with the conventional treatment. A
report from the ANCS0332 randomized trial focussed on the molecular
heterogeneity within the umbrella term cPNET and showed that the
outcomes of these patients were considerably better when histologically
cPNET, but molecularly high-grade gliomas (HGGs) were excluded from the
analysis using DNA methylation profiling(29). This trial also showed
that unlike high-risk MB patients, cPNET patients did not derive any
event-free survival (EFS) benefit with the use of carboplatin
concurrently with radiotherapy, so vincristine is sufficient for cPNET
patients.
Non-PB cPNET was found to be resistant to Packer’s chemotherapy
regimen(30). Intensive chemotherapy without radiotherapy was found to
jeopardize the survival, and so was the omission of CSI(31). A report of
the Head Start I and II trials experience, however, showed that
postoperative intensified induction chemotherapy followed by
myeloablative chemotherapy and autologous stem cell transplant, without
irradiation was associated with better outcomes and avoidance of
CSI-associated adverse effects. RT was reserved for salvage setting;
local recurrences were much higher than local and distal recurrences and
60% patients alive at 5 years had no RT exposure. However, long-term
data are not available(32). Massimino et al., after debulking surgery,
used high-dose chemotherapy (methotrexate, etoposide, cyclophosphamide
and carboplatin ± vincristine), followed by hyperfractionated
accelerated RT, two daily 1.3 Gy fractions, to a dose of 31.2Gy and 39Gy
in less than 10 and more than 10 years old patients respectively. Local
boost was delivered in two daily 1.5 Gy fractions, up to 59.7-60 Gy. RT
was followed by myeloablative dose of thiotepa and autologous stem cell
rescue. Upon the observation that local failure was seen even after CSI
and that no isolated distal relapse occurred in first 15 patients
treated, local conventionally fractionated RT to a dose of 54 Gy was
attempted in rest of the patients with localized disease and with no
progression during induction chemotherapy. The results of focal RT,
analysed separately were better than the entire series(33).
Chintagumpala et al. classified cPNET patients into average-risk
(M0 and residual tumor < 1.5
cm2) and high-risk (neuraxial dissemination or
residual > 1.5 cm2) and planned adjuvant RT as per MB
guidelines for each risk group, i.e., lower CSI dose for average risk.
RT was followed by 4 cycles of nonmyeloablative high-dose chemotherapy
(high-dose cyclophosphamide, cisplatin, and vincristine), each cycle
with stem cell support. Average-risk cPNET treated with lower CSI dose
and high-dose chemotherapy with stem cell rescue had excellent 5-year
EFS of 75%, thereby highlighting the advantage of risk-adapted
approach(34). Timmerman et al analysed the results of two trials, done
on less than 3 years old children with supratentorial PNET to compare
intensive postoperative chemotherapy alone and adjuvant induction
chemotherapy followed by delayed RT. They concluded that RT should not
be omitted despite intensive chemotherapy and even in less than 3 years
old, RT should not be delayed for more than 6 months(31).
Our first patient had a residual tumour more than
1.5cm2 post-surgery without any evidence of CSF
dissemination, and was treated as per high-risk MB protocol, though
vincristine, and not carboplatin was used concurrently with RT(29). The
adjuvant chemotherapy planned was Packer’s A regimen. However, the
second patient had a smaller residual (0.8 cm2)
post-surgery, without any CSF dissemination. She was also treated as per
high-risk MB protocol with vincristine concurrent with RT and adjuvant
chemotherapy as per Packers A regimen.