- Email: [email protected]
Presentation of childhood CNS tumours: a systematic review and meta-analysis
Articles
Presentation of childhood CNS tumours: a systematic review and meta-analysis Sophie Wilne, Jacqueline Collier, Colin Kennedy, Karin Koller, Richard Grundy, David Walker
Summary Background Suspicion of a CNS tumour is classically raised by symptoms of raised intracranial pressure, focal deficits (including seizures), or papilloedema. Development of guidelines is needed for the identification and referral of children who might have a CNS tumour. We did a systematic literature review and meta-analysis to identify the clinical presentation of childhood CNS tumours to provide evidence to support the development of guidelines to assist with the identification and referral for imaging of children who might have a central nervous system tumour. Methods Medline, Embase, and PubMed were searched for cohort studies and case series in children, published between January, 1991, and August, 2005, detailing the symptoms and signs at diagnosis of a CNS tumour. Findings 74 papers (n=4171) met the inclusion criteria. 56 symptoms and signs at diagnosis were identified, ranked by frequency, and clustered according to age, anatomical criteria, and genetic criteria. The most frequent symptoms and signs at diagnosis were: headache (33%), nausea and vomiting (32%), abnormalities of gait and coordination (27%), and papilloedema (13%) for intracranial tumours; macrocephaly (41%), nausea and vomiting (30%), irritability (24%), and lethargy (21%) for children aged under 4 years with intracranial tumours; reduced visual acuity (41%), exophthalmia (16%), and optic atrophy (15%) for children with an intracranial tumour and neurofibromatosis; nausea and vomiting (75%), headache (67%), abnormal gait and coordination (60%), and papilloedema (34%) for posterior fossa tumours; unspecified symptoms and signs of raised intracranial pressure (47%), seizures (38%), and papilloedema (21%) for supratentorial tumours; headache (49%), abnormal eye movements (21%), squint (21%), and nausea and vomiting (19%) for central brain tumours; abnormal gait and coordination (78%), cranial nerve palsies (52%), pyramidal signs (33%), headache (23%), and squint (19%) for brainstem tumours; and back pain (67%), abnormalities of gait and coordination (42%), spinal deformity (39%), focal weakness (21%), and sphincter disturbance (20%) for spinal-cord tumours. Other features noted were weight loss, growth failure, and precocious puberty. Symptoms of raised intracranial pressure were absent in more than half of children with brain tumours. Other neurological features were heterogeneous and related to tumour location. Interpretation Apart from raised intracranial pressure, motor and visual system abnormalities, weight loss, macrocephaly, growth failure, and precocious puberty also suggest presence of an intracranial tumour. Children with signs and symptoms that could result from a CNS tumour need a thorough visual and motor system examination and an assessment of growth and pubertal status. Occurrence of multiple symptoms and signs should alert clinicians to possible CNS tumours.
Introduction Life-threatening clinical conditions in childhood are seen infrequently in developed countries.1,2 Screening of the few serious diagnoses from the many self-limiting conditions and fluctuations in developmental processes and behaviour is a major diagnostic challenge in primary and secondary health care.3,4 Cancer affects one in 600 children under 16 years and thus represents a moderate health risk similar to cerebral palsy, diabetes mellitus, and meningitis.5–7 A quarter of childhood cancers arise in the CNS and account for the largest number of cancer deaths in childhood.1 60% of survivors are left with pronounced disability.8–12 CNS tumours are therefore common in the context of life-threatening childhood disease. All healthcare professionals seeing children should be able to identify symptoms and signs that could result from a CNS tumour and refer or investigate appropriately. Despite advances in neuroimaging, the timely diagnosis of CNS tumours in the UK remains difficult.13 http://oncology.thelancet.com Vol 8 August 2007
Lancet Oncol 2007; 8: 685–95 Published Online July 18, 2007 DOI:10.1016/S14702045(07)70207-3 See Reflection and Reaction page 664 Children’s Brain Tumour Research Centre, Academic Division of Child Health (S Wilne MBBS, K Koller PhD, Prof R Grundy MBChB, Prof D Walker MBBS), and School of Nursing, Faculty of Medicine & Health Sciences (Prof J Collier PhD), University of Nottingham, Queen’s Medical Centre, Nottingham; and Department of Child Health, University of Southampton, Southampton General Hospital, Southampton, UK (Prof C Kennedy MBBS) Correspondence to: Dr Sophie Wilne, Children’s Brain Tumour Research Centre, Academic Division of Child Health, University of Nottingham, Queens Medical Centre, Nottingham NG7 2UH, UK sophie.wilne@nottingham. ac.uk
The varied presentation and perceived rarity (resulting in a low priority in the differential diagnosis) of CNS tumours in childhood underlie this problem. Many of the initial symptoms and signs of CNS tumours also occur with other more common and less serious childhood disorders such as gastroenteritis, migraine, and behavioural problems. Conventional teaching is that CNS tumours present with symptoms of raised intracranial pressure (early morning headache with vomiting and papilloedema) with or without focal neurological signs.14,15 Research in the pre-CT era has identified comprehensive lists of symptom and sign clusters with which neurologists and other paediatric specialists are highly familiar. Currently, expanded access to neuroimaging and the pressure to accelerate cancer diagnoses is placing a much broader group of doctors in the position to initiate imaging either directly or through an urgent cancer referral. A period of diagnostic uncertainty often precedes the diagnosis of a 685
Articles
Data collection MEDLINE, PubMed, and EMBASE searched from January, 1991, to August, 2005 5620 potentially relevant studies identified and abstracts screened for retrieval
5234 abstracts excluded because subject not related to paediatric CNS tumour presentation
386 papers retrieved for detailed review
312 papers excluded 32 had combined adult and paediatric data 4 had no paediatric data 9 had less than ten children 20 had insufficient detail on signs and symptoms 7 had combined CNS and extradural tumours 4 had no primary data 236 had unrelated subject matter or duplicate study
74 papers included in meta-analysis
Figure 1: Progress through the meta-analysis
CNS tumour, which patients and their families find extremely distressing, and any perception that the medical response has been inadequate, incompetent, or delayed may be associated with legal dispute.16 In 1991, the Childhood Brain Tumour Consortium17 reported the symptoms and signs at diagnosis for 3291 children diagnosed with a brain tumour in 1930–79. To provide a more contemporary cohort, we undertook a systematic review and meta-analysis of the presenting symptoms and signs in paediatric CNS tumours detailed in subsequent publications. The literature review was done as the initial stage in a project devising guidance to help health-care professionals identify children who might have a CNS tumour and who thus need fast-track imaging.
Methods Identification of studies and inclusion criteria We searched MEDLINE, PubMed, and EMBASE without language restriction, from January, 1991, to August, 2005. Key words were: “brain tumour(s), “brain tumor(s)”, “brain neoplasm(s)”, “spinal cord tumour(s)”, “spinal cord tumor(s)”, “spinal cord neoplasm”; and “diagnosis”; and “sign(s)” or “symptom(s)”. Retrieved references were restricted to “all child”. Abstracts were screened by a researcher (SW); those unrelated to CNS tumours or discussing an area unrelated to clinical presentation were excluded. Papers with abstracts discussing tumour presentation, tumour diagnosis, or clinical symptoms and signs were retrieved for detailed review. We included all case-series or cohort studies describing symptoms and signs at diagnosis for a minimum of ten children diagnosed with a CNS tumour and published after February, 1991. Non-English language papers were translated. 686
Numbers of children in every study with a symptom or sign at diagnosis were recorded on a standard data extraction form. Information on symptoms and signs varied between studies. Some studies had very detailed records on individual symptoms and signs (eg, headache, vomiting, papilloedema), whereas others reported symptoms in clusters or complexes (eg, symptoms of raised intracranial pressure). Symptoms and signs were recorded as described in the individual studies. If a symptom or sign was not recorded in a study, we assumed it not to occur in that population.
Statistical analysis Analysis was done with meta-disc version β 1.1.1. We combined proportions (%) of children with each symptom or sign at diagnosis using one-variable relationship metaanalysis. The effect size for each symptom and sign was calculated in the individual studies and weighted according to its variance, and these effect sizes were then summed (for each symptom and sign) and the total effect size was then divided by the sum of the weights to give a mean effect size (pooled proportion).18 In meta-disc, proportions (as well as likelihood ratios and diagnostic ratios) could be pooled with either the Mantel-Haenszel method (fixed-effects model) or, to incorporate variation between studies, with the DerSimonian Laird method (random-effects model). In the present study, heterogeneity was indicated beyond what could be expected by chance alone, by significant Q statistics and high inconsistency (I²) statistics. The DerSimonian Laird method was selected because variability was expected across the papers, and a randomeffects model was used.18 Symptoms and signs occurring in 5% or more of the meta-analysis population were reported. Two papers19,20 reported optic atrophy and papilloedema and one paper21 lethargy and irritability as a combined category. Since these papers reported detailed information for other symptoms and signs, they were included in the meta-analysis but excluded from the analysis of the combined symptoms or signs. In one report,19 visual acuity was not assessed in the complete cohort and, therefore, was excluded from the metaanalysis of visual acuity. The following subgroup analyses were undertaken: all intracranial tumours; intracranial tumours in children aged under 4 years; children with an intracranial tumour and neurofibromatosis; posterior fossa tumours; supratentorial (excluding central) tumours; central tumours (third ventricle, tectum, pineal gland, pituitary gland, thalamus, hypothalamus, optic pathway, and basal ganglia); brainstem tumours; and spinal-cord tumours. Analysis of all intracranial tumours was undertaken to provide a summary of paediatric intracranial tumour presentation. Children aged under 4 years usually cannot clearly describe symptoms such as headache, nausea, and diplopia, and therefore have a different presentation to http://oncology.thelancet.com Vol 8 August 2007
Articles
Recruitment Number of period patients
Patient group, diagnosis if known, source of data
Tumour location
Mean age, years
Median age, years
Age range, years
Mean symptom Median symptom Symptom Ref interval, interval, interval range, months months months NS
1977–87
22
Infants, 1I*
AB
NS
NS
NS
1981–89
16
Under 2 years, 1I*
AB
NS
NS
NS
NS
1965–89
16
NF1 and OP tumours, 2I
OP
NS
4·5
1·5–17
1978–91
12
Gangliogliomas, 1I
AB
NS
1977–90
12
Gangliogliomas, 1I
AB
1976–91
12
Midbrain tumours, 1I
C
1975–81
11
Choroid plexus carcinoma, 2I
AB
1976–88
21
Meningeal tumours, 1I
AB
NS
0·3–16·7
14·6
1962–89
39
Under 2 years, 1I*
AB
NS
NS
NS
NS
NS
NS
30
1970–89
106
Cerebral hemisphere tumours, 1I
ST
NS
NS
NS
NS
NS
NS
31
1970–87
80
Under 2 years at symptom onset, 1I*
AB
NS
1989–92
14
Infants with supratentorial tumours, 1I
ST
0·5
1980–90
10
Meningiomas, 1I
AB
1973–92
21
NF1 and OP tumours, 4I
OP
1979–94
21
Under 2 years, 1I*
AB
1983–92
17
Midbrain tumours, 1I
C
1974–94
23
Intracranial ependymoma, 1I
AB
1984–94
17
NF1 and BS tumours, 1I
BS
1990–94
74
AB tumours, 1I
AB
1988–91
119
BS gliomas treated with hyperfractionated radiotherapy (Children’s Cancer Group trial 9882)
BS
2·5
1
25·2
NS
22
0·5–9
23
NS
NS
24
NS
3·5–17
NS
NS
NS
25
7·8
NS
0·8–15·8
40
27
7–96
26
8·2
NS
1·1–16
2·2
0–9·5
NS 9·3
4·5 NS
NS
0·5–14
27
NS
NS
28
0–72
29
4
NS
NS
NS
NS
0–153·6
21
NS
0·1–0·9
NS
NS
NS
20
11·1
NS
8–15
13·2
NS
0·1–60
32
7·1
NS
0–14·5
NS
NS
NS
33
NS
NS
0·2–1·8
NS
NS
NS
34
NS
9·7
3·5–16
NS
4
NS
35
8·8
NS
2–14
8·4
8·3
1·3–13·9
6·9
NS
NS
6·5
NS
NS
NS
3·8 NS 4·6
NS
0·5–10
36
NS
NS
37
NS
0·2–30
38
NS
NS
39
1984–93
32
Gangliogliomas, 1I
AB
6·5
NS
0·7–20
NS
NS
NS
40
1970–95
36
Supratentorial primitive neuroectodermal tumours, 1I
ST
4·3
2·9
0·1–12·8
NS
NS
NS
41
1980–93
27
Under 3 years with intramedullary SC, 1I
SC
1·7
NS
0·5–3
NS
NS
NS
42
1984–95
13
Intrinsic SC tumours, 1I
SC
5·4
NS
0·7–11
NS
NS
NS
43
1984–95
723
AB tumours, 1I
AB
NS
NS
0–16
NS
NS
NS
44
1980–90
35
BS tumours, 1I
BS
NS
NS
1·3–13
5
NS
NS
45
1987–94
30
Endophytic pons or medullary tumours, 1I
BS
NS
6
0·6–16
6
NS
1–60
46
1974–95
99
Gangliogliomas, 1I
AB
NS
1·7–20
60
24·4
NS
47
1968–94
29
Meningiomas 2I
AB
10
NS
0–15
NS
NS
NS
48
1983–95
12
Primary intracranial germ-cell tumours, 1I
C
NS
NS
5–15
NS
NS
NS
49
1984–96
25
NF1 and BS tumours, 1I
BS
7·8
NS
1·1–15·2
NS
NS
NS
50
1976–92
18
SP astrocytomas, 1I
SC
9·2
8·6
0·6–17·9
NS
NS
NS
51
1966–96
46
Under 3 years, 1I*
AB
NS
NS
NS
NS
NS
52
1985–95
20
Under 3 years, 1I*
AB
1·7
NS
0–2·7
1977–96
21
Infants, 1I
AB
0·5
NS
NS
NS
1990–97
32
Tectal tumours, 1I
C
8
NS
0·2–17
1984–95
22
Choroid plexus carcinoma registered with Societe Francais Oncologie Paediatric
AB
NS
2·1
0·3–9·3
1986–90
40
Intracranial ependymoma treated in Paediatric Oncology Group trial 8633
AB
NS
NS
0·3–2·9
1971–94
73
SP astrocytomas, 13I
SC
NS
7
0·3–6
NS
NS
NS
58
1985–96
20
Intramedullary SP ependymomas, 1I
SC
14
NS
9–18
NS
NS
NS
59
1975–93
200
All brain tumours, 1I
AB
8·9
NS
NS
NS
NS
NS
60
1987–97
39
Under 3 years, 1I*
AB
2·1
NS
0·3–3
NS
0·2–18
61
1983–97
76
BS gliomas
BS
NS
NS
3–15
NS
NS
NS
62 63
9·5
NS
2·4
0–18
53
NS
NS
54
NS
NS
NS
55
NS
1
0·1–8
56
1
0–10·9
57
1·6
5·2
1
1988–98
11
Tectal-plate gliomas, 1I
C
10
NS
5–13
28·2
NS
0·7–84
1988–98
54
Lateral-ventricle tumours, 1I
ST
NS
NS
0–20
5
NS
0–48
64
1967–97
37
Pineal-region tumours, 1I
C
9·6
NS
NS
NS
NS
NS
65
1986–95
28
Supratentorial primitive neuroectodermal tumours, 1I
ST
6·8
NS
0·7–16·9
1988–98
11
Cervicomedullary astrocytomas, 1I
SC
7
NS
0–18
4·9 NS
NS
1–48
66
NS
NS
67
(Continues on next page)
http://oncology.thelancet.com Vol 8 August 2007
687
Articles
(Continued from previous page) 1984–95
22
Any brain tumours reported to regional tumour registry
AB
NS
1979–99
34
Choroid plexus tumours, 1I
AB
NS
1·4
0·1–11·5
NS
1
1972–91
62
Intracranial ependymoma, 1I
PF
6
NS
1–17
2
NS
1984–99
24
Meningiomas, 2I
AB
NS
NS
2–17
8·2
NS
0·2–14·4
71
1980–94
18
Chiasmal gliomas, 1I
OP
NS
NS
0·5–14
NS
NS
NS
72
1985–99
181
AB tumours, 1I
AB
NS
NS
0–16
NS
NS
NS
73
1970–98
16
Choroid plexus tumours, 1I
AB
NS
0·2–15·4
NS
NS
NS
74
1974–99
122
Medulloblastoma, 1I
PF
NS
NS
NS
NS
NS
75
1981–98
11
Nerve-cell tumours, 1I
ST
NS
NS
2–16
NS
NS
NS
76
1970–98
35
Craniopharyngiomas, 1I
C
NS
9·1
1·3–15·6
NS
NS
NS
77
1980–99
252
AB, 1I
AB
NS
NS
1995–2000
104
AB, 2I
AB
1987–99
22
Gangliogliomas, 2I
AB
1980–2000
20
Thalamic and basal ganglia tumours, 1I
C
1974–99
18
Meningiomas recorded in a hospital tumour registry
AB
1975–2002
61
Infants, 2I*
AB
1988–99
16
Infants, 1I*
AB
1986–90
13
Supratentorial primative neuroectodermal tumours treated in Pediatric Oncology Group trial 8633
1954–97
181
1982–2000 1996–2000
69
3·1
8·29 NS 6·6
NS
6·3
0–16·9
NS
NS
NS
0–16
NS
0·3–18
4·6
3·3
7·3 30 1·5
0·2–45·9
68
0·03–33
69
NS
70
1·8
0–99
78
3
NS
79
11
NS
80
NS
0–24
81
NS
NS
82
NS
0·1–8·9
83
0.2
0–6
84
NS
0–49
85
1·6–17
NS
0–1
NS
0·5
0–1
ST
NS
NS
0–3
Medulloblastoma registered with Manchester Children’s tumour registry
PF
NS
NS
0–14
NS
NS
NS
86
NF1 and symptomatic tumours, 7I
AB
NS
5·2
0·3–17
NS
NS
NS
87
OP
NS
NS
0·3–17·4
NS
NS
NS
19
NS
NS
NS
88
NS
NS
89
2·5–96
90
0·5
1986–2002
51
NF1 and symptomatic optic pathway gliomas, 2I
OP
4·8
NS
0–15·8
1996–2003
37
PF tumours, 1I
PF
6·7
NS
2–16
1978–2001
18
Giant-cell astrocytomas, 2I
AB
1973–2002
57
NF1 and OP tumours, 1I
OP
1997–2000
172
AB, 1I
AB
NS 5·2 NS
NS
2·1
NS
83 (51 NF1) OP gliomas, 2I
11
NS
1·9 NS 0·9
3·7
NS
4–15
19
9
NS
NS
NS
NS
8·3
0·3–17·3
4·9
1
NS
91
0·2–120
92
1I=treated at one institution. 2I=treated at two institutions. 4I=treated at four institutions. 7I=treated at seven institutions. AB=all brain. NS=not specified. OP=optic pathway. C=central. ST=supratentorial. BS=brainstem. SC=spinal cord. PF=posterior fossa. *Study population defined by age rather than tumour type or location.
Table: Studies meeting inclusion criteria
older children. Neurofibromatosis is the commonest genetic abnormality associated with intracranial tumours and children can develop tumours before the development of cutaneous manifestations. Children with neurofibromatosis have a high occurrence of optic-pathway tumours, and thus their presentation differs from that of other children with intracranial tumours. Only children with neurofibromatosis and a symptomatic intracranial tumour were included in this subgroup analysis. We did not analyse asymptomatic children with an intracranial tumour identified by CNS imaging that was instigated after a diagnosis of neurofibromatosis. Analysis by tumour location was undertaken to highlight specific associations of symptoms and signs that occur with different tumour locations.
Role of the funding source The sponsors of the study had no role in study design, data collection, data analysis, or data interpretation, in the writing of this report, or in the decision to submit the paper for publication. All authors had full access to all 688
the data in the study, and the corresponding author had final responsibility for the decision to submit for publication.
Results 5620 papers were identified by the search terms. We reviewed 386 papers in full, from which 74 met inclusion criteria, describing the symptoms and signs at diagnosis in 4171 children (figure 1, table).19–92 56 symptoms and signs were recorded in children with CNS tumours, but only symptoms and signs that occurred in 5% or more of patients are reported here. 61 studies (n=3702)19–23,25–32,34–36,38–41,44–49,52–57,60–66,68–86,89,90,92 described the symptoms and signs at diagnosis for children without neurofibromatosis who had an intracranial tumour. These were (in decreasing order of frequency): headache (33%), nausea and vomiting (32%), abnormal gait or coordination (27%), papilloedema (13%), seizures (13%), unspecified symptoms and signs of raised intracranial pressure (10%), squint (7%), change in behavioural or school performance (7%), macrocephaly (7%), cranial http://oncology.thelancet.com Vol 8 August 2007
Articles
nerve palsies (unspecified; 7%), lethargy (6%), abnormal eye movements (nystagmus, Parinaud’s syndrome; 6%), hemiplegia (6%), weight loss (5%), focal motor weakness (5%), unspecified visual or eye abnormalities (5%), and altered level of consciousness (5%; figure 2). 13 studies (n=332)20–23,30,34,52–54,57,61,83,84 were included in the analysis of children with intracranial tumours aged under 4 years. Ranked symptoms and signs at diagnosis were: macrocephaly (41%), nausea and vomiting (30%), irritability (24%), lethargy (21%), abnormal gait and coordination difficulties (19%), weight loss (14%), clinically apparent hydrocephalus (bulging fontanelle, splayed sutures; 13%), seizures (10%), papilloedema (10%), headache (10%), unspecified focal neurological signs (10%), unspecified symptoms of raised intracranial pressure (9%), focal motor weakness (7%), head tilt (7%), altered level of consciousness (7%), squint (6%), abnormal eye movements (6%), developmental delay (5%), and hemiplegia (5%; figure 2) Eight studies (n=307)19,24,33,37,50,87,89,92 were included in the analysis of children with neurofibromatosis and an intracranial tumour. The most common symptom and signs at diagnosis were visual, indicating the high occurrence of optic pathway gliomas in this population. The ranked symptoms and signs were reduced visual acuity (41%), exophthalmia (16%), optic atrophy (15%), squint (13%), headache (9%), unspecified symptoms of raised intracranial pressure (8%), precocious puberty (8%), abnormal gait or coordination difficulties (7%), voice abnormalities (6%), developmental delay (5%), papilloedema (5%), and reduced visual fields (5%; figure 2). Five studies (n=476)60,70,75,86,89 described children with posterior fossa tumours; seven studies (n=303)20,31,41,60,64,66,85 described children with supratentorial tumours; 11 (n=276)19,27,35,49,55,60,63,65,72,77,81 described children with central tumours; five (n=276)39,45,46,60,62 described children with brainstem tumours; and six studies (n=162)42,51,58,59,67,74 described children with spinal-cord tumours (figures 3 and 4).
Discussion To our knowledge, our study is the first systematic review and meta-analysis of symptom and sign patterns in children with CNS tumours. Our findings show the importance of age, tumour location, and neurofibromatosis status in the determination of the symptom and sign clusters present at tumour diagnosis. Combining the most common specific symptom or sign of raised intracranial pressure with the proportion of children presenting with non-specific symptoms or signs of raised intracranial pressure provides an estimate of the overall frequency of these symptoms and signs. This relation indicates that symptoms linked to raised intracranial pressure are present in about 40% of all intracranial tumours, 40% of intracranial tumours in children aged under 4 years, 20% of intracranial tumours occurring in http://oncology.thelancet.com Vol 8 August 2007
All intracranial tumours
Children under 4 years
Children with neurofibromatosis
Headache Nausea and vomiting Papilloedema Macrocephaly Clinically apparent hydrocephalus Symptoms of raised ICP NOS Abnormal gait or coordination Hemiplegia Focal motor weakness Head tilt Squint Abnormal eye movements Reduced visual acuity Reduced visual fields Optic atrophy Exophthalmia Visual or eye signs NOS Impaired consciousness Lethargy Behavioural or school change Irritability Weight loss Developmental delay Voice abnormalities CNP NOS Seizures Precocious puberty Focal neurological signs NOS Pooled proportion 95% CI
0
10 20 30 40 50 0
10 20 30 40 50 0
10 20 30 40 50
Pooled proportion of children affected (%)
Figure 2: Frequency of symptoms and signs in children with intracranial tumours Analysis by age and neurofibromatosis status. ICP=intracranial pressure. NOS=not otherwise specified. CNP=cranial nerve palsy.
children with neurofibromatosis, 80% of posterior fossa tumours, 60% of central tumours, 60% of hemispheric tumours, 30% of brainstem tumours, and 7% of spinalcord tumours. Other alerts to a possible CNS tumour include abnormal gait and coordination, other motor system abnormalities, eye signs, weight loss, behavioural changes (including lethargy and irritability) and school difficulties, developmental delay, cranial nerve palsies, head tilt, macrocephaly, diabetes insipidus, and growth arrest. Increasing awareness of the varied and complex symptomatology that often occurs with CNS tumours could help tumour diagnosis and reduce the extended symptom interval experienced by many children. Recognition that specific combinations of symptoms and signs indicate a focal CNS lesion is crucial to the diagnosis of many CNS tumours. 45–60% of childhood brain tumours are infratentorial, 25–40% are hemispheric, and 15–20% are midline supratentorial.93 Our approach has emphasised the symptom and sign combinations that occur with different tumour locations and will help 689
690
95% CI
Pooled proportion
Sphincter disturbance
Spinal deformity
Back pain
Accidental head injury
Focal neurological signs NOS
Short stature
Diabetes insipidus
Seizures
Cranial nerve palsies
Auditory symptoms or vertigo
Facial palsy
Developmental delay
Weight loss
Behavioural or school change
Lethargy
Altered level of conciousness
Visual or eye signs NOS
Optic atrophy
Reduced visual fields
Reduced visual acuity
Abnormal eye movements
Squint
Stiff neck
Decreased upper limb movement
Head tilt
Focal motor weakness
Pyramidal signs
Hemiplegia
Abnormal gait or coordination
Symptoms of raised ICP (NOS)
Macrocephaly
Papilloedema
Nausea without vomiting
Nausea and vomiting
Headache
0
10 20 30 40
50 60 70 80 90 100 0
Posterior fossa tumours
10 20 30 40
50 60 70 80 90 100 0
Supratentorial (excluding central) tumours
50 60 70 80 90 100 0 Pooled proportion of children affected (%)
10 20 30 40
Central tumours
10 20 30 40
50 60 70 80 90 100 0
Brain-stem tumours
10 20 30 40
50 60 70 80 90 100
Spinal-cord tumours
Articles
Figure 3: Frequency of symptoms and signs in children with CNS tumours Analysis by tumour location. ICP=intracranial pressure. NOS=not otherwise specified.
http://oncology.thelancet.com Vol 8 August 2007
Articles
Brain stem tumours:
Supratentorial tumours: Unspecified symptoms of raised ICP* 47% Seizures 38% Papilloedema* 21% Focal neurological signs 17% Headache* 11% Hemiplegia 10% Nausea and vomiting* 8% Macrocephaly* 6%
Abnormal gait and coordination difficulties 78% Cranial nerve palsies (unspecified) 52% Pyramidal signs (unspecified) 33% Headache* 23% Squint 19% Focal motor weakness 19% Facial palsy 15% Papilloedema* 13% Unspecified symptoms of raised ICP* 10% Abnormal eye movements 6% Behavioural change or school difficulties 5%
Posterior fossa tumours: Nausea and vomiting* 75% Headache* 67% Abnormal gait and coordination difficulties 60% Papilloedema* 34% Abnormal eye movements 20% Lethargy 13% Nausea without vomiting* 10% Unspecified symptoms and signs of raised ICP* 9% Weight loss 9% Focal motor weakness 9% Macrocephaly* 7% Impaired consciousness 7% Vertigo or auditory symptoms 7% Squint 6% Stiff neck 6% Head tilt Accidental head injury 5%
Spinal cord tumours: Back pain 67% Abnormal gait or coordination difficulties 42% Spinal deformity 39% Focal motor weakness 21% Sphincter disturbance 20% Decreased upper limb movement 17% Developmental delay 8% Head tilt 7% Headache* 7%
Central tumours: Headache* 49% Abnormal eye movements and squint 21% Nausea and vomiting* 19% Papilloedema* 18% Reduced visual acuity 16% Unspecified symptoms and signs of raised ICP* 13% Diabetes insipidus 12% Abnormal gait and coordination difficulties 10% Optic atrophy 9% Behavioural change or school difficulties 9% Altered level of consciousness 9% Reduced visual fields 8% Seizures 7% Hemiplegia 7% Focal motor deficit 7% Developmental delay 7% Short stature 7% Weight loss 5% Vertigo or auditory symptoms 5% Visual or eye abnormalities (unspecified) 5%
Figure 4: CNS tumour presentation *Symptom or sign caused by raised intracranial pressure (ICP).
focus the search for corroborative findings in children who present with a symptom or sign that is potentially suggestive of a CNS tumour. In many instances, the possibility that the symptoms or signs are the result of a CNS tumour will be (rightly) rapidly dismissed. However, consideration of this diagnosis in some cases could lead to identification of corroborative symptoms and signs and the instigation of imaging. Even if an underlying tumour is unlikely, patients and their families or carers should be encouraged to return for re-assessment should symptoms or signs persist or progress, and the diagnosis should be reviewed on re-presentation. We chose a 5% threshold for reporting symptoms and signs in children with CNS tumours as a practical compromise between the need to consider an underlying CNS tumour with a clinical feature not associated with this tumour type and those symptoms and signs that occur frequently in childhood CNS tumours. Because of the differing presentation of CNS tumours according to patient group and tumour location, most symptoms and http://oncology.thelancet.com Vol 8 August 2007
signs that occurred in less than 5% of patients in one subgroup occurred more frequently in the other subgroups. Symptoms and signs that consistently occurred in less than 5% of patients, which could be associated with diagnostic difficulty, were dysphagia and delayed puberty. 35 studies21,23,24,26,27,29,32,35,36,38,45-47,53,56,57,61,63,64,66,68–71,75,78–81,83–85,89,90,92 meeting the inclusion criteria reported symptom interval duration (table). Symptom interval comparison is difficult for several reasons. Studies report different measures of symptom interval (median, mean, range) and rarely report all three, and the statistical significance of any differences in symptom interval cannot be determined from the reported data. For asymmetric distributions such as symptom interval of paediatric brain tumours, the median provides the best comparator. The reported median symptom interval ranges from 1 to 27 months. The longest median symptom interval occurs with biologically slowgrowing tumours such as gangliogliomas,26,47 although there is little association between symptom interval range 691
Articles
and tumour biology, indicating that extended symptom intervals could occur with all types of paediatric brain tumour. Any association between specific symptoms and signs and an extended symptom interval could not be identified in this study; exploration of any such relation would need a large cohort study in which the symptom interval and time of onset for individual symptoms and signs was recorded for all patients. We used a systematic search strategy and standardised inclusion criteria, as recommended in the quality of reporting meta-analyses (QUOROM) statement, to identify studies for inclusion.94 The high number of papers identified in the past 15 years shows the sustained interest in the mechanisms of diagnosis in this group of patients. Previously published evidence on paediatric CNS tumour presentation has been predominantly in the form of case studies (level 4 evidence) with infrequent tumour registry series (level 2 evidence; two studies met the inclusion criteria for this study).68,86,95 The systematic approach has generated a cohort of patients, most of whom were diagnosed during the era of CT and MRI, six times larger that the largest single identified study. The meta-analysis results reported here provide level 2 evidence for this cohort, which give greater value to the rankings of symptoms and signs by age, tumour location, and neurofibromatosis status than previous reports. This study had some important limitations and potential sources of bias. The search strategy might not have identified all relevant papers and unpublished data were not sought. Papers included in the analysis reported symptoms and signs at diagnosis in children with a CNS tumour; therefore accuracy of these data depends on the history given by patients and their families or carers and the signs detected by the examining health-care practitioners. However, medical decisions will always be based on such histories and examination findings rather than the underlying full facts to which they relate. The assumption was made that if a symptom or sign was not described in a study, it did not occur in that population. The variability and large number of included patients should reduce the risk that common symptoms and signs are underrepresented and uncommon ones over-represented. We saw variation in the data detail between studies. Some studies were very detailed, recording individual symptoms and signs such as headache, vomiting, and papilloedema; 27–29,32,34–36,38,40,44,47,53,55,57,60–63,65,66,68,70,73,75–82,85,86,89,92 whereas others used symptom complexes such as symptoms of raised intracranial pressure or cranial nerve palsies.19,31,39,41,48,52,56,60,64,69,71,83 Some symptoms and signs could have been combined to indicate the total proportion of children presenting with a specific symptom complex. However, since we could not clearly deduce exactly how the data related, some inaccuracy and misrepresentation of data could result and thus we kept data in their original form. Despite these problems, the analysis shows the variability of symptoms and signs and the frequency with which they occur in childhood CNS tumours. 692
Most childhood brain tumours are low-grade astrocytomas.96,97 Apart from optic pathway gliomas, these astrocytomas were under-represented in the studies identified. This result is probably due to a historical failure to include non-malignant brain tumours in tumour registries and, until recently, absence of review of children with low-grade gliomas by paediatric oncologists. Despite this result, the distribution of tumour location in the studies identified here was similar to that seen in clinical practice (56% infratentorial, 23% hemispheric, 21% central), lending support to the analysis results. Publication bias could have led to overrepresentation of rare tumours or those with an unusual presentation; however, we excluded case reports and studies with fewer than ten patients to combat this problem. Finally, this analysis addresses the issue of sensitivity but not that of specificity of symptoms and signs to the presence of an underlying CNS tumour. Therefore, our findings might be more useful in guiding clinical-management decisions in children who are already selected by referral to secondary or tertiary care than in unselected children. The probability of a symptom or sign being indicative of a CNS tumour will increase with the occurrence of corroborative findings on history and examination and the prevalence of CNS tumours in the population in question. The previous largest study17 of childhood brain tumour presentation, undertaken by the Childhood Brain Tumor Consortium, reported on the distribution of other symptoms and neurological signs in 3291 children with or without headache in association with a brain tumour. For the most of this period, CT and MRI were not available. Direct comparison between this study17 and the present analysis is complicated by differences in anatomical subdivision and methodology. Headache, nausea and vomiting, and seizures were reported for the entire cohort, although other symptoms were reported for specific age groups and numbers in each age group were not provided.17 Similarly, although the occurrence of coma, focal motor weakness, and papilloedema is reported for the entire group, other symptoms were not reported unless their presence or absence was documented in the medical records. Notably, the Childhood Brain Tumor Consortium cohort17 reported a higher frequency of headache, nausea and vomiting, and papilloedema in supratentorial tumours than our study, but reported a similar frequency of these symptoms in infratentorial tumours. This difference is probably due to increased imaging availability to our cohort. Because of the vulnerability of the cerebral aqueduct to compression by tumour, posterior fossa tumours often lead to raised intracranial pressure at an early stage. By contrast, supratentorial tumours could present with other symptoms and signs and grow to a large size before they lead to raised intracranial pressure. The availability of CT and MRI allows the latter children with supratentorial tumours to be assessed before the development of raised intracranial pressure. http://oncology.thelancet.com Vol 8 August 2007
Articles
In the meta-analysis, the frequency of change in behaviour or school performance (7% in all brain tumours, and 9% for central tumours) was lower than that reported by many individual studies. Several large cohorts reported a frequency of school difficulties and behavioural changes of 22–72%.38,44,68,98 Lethargy was analysed separately in this study (pooled proportion: 6% for all intracranial tumours, 21% in children with intracranial tumours aged under 4 years, 13% in posterior fossa tumours), which could account for some of the difference. Adults with brain tumours are often not asked about behavioural change, and similar reporting errors probably occur in children.99 The literature search for this analysis identified several deficits in this specialty. Most identified studies were done in one centre. Only four studies reported symptoms and signs from children enrolled in national trials, of which two reported low numbers of patients.39,56,57,85 Many multi-institutional and multinational trials are done in paediatric neuro-oncology, although these studies, while reporting survival, rarely report symptoms and signs. These symptoms and signs would be easy to obtain, and would improve the level of evidence in this area. Patients and their families find the extended symptom interval associated with paediatric brain tumours very distressing,16 and would probably be willing to provide this information if it could aid earlier diagnosis for future children. The increasing involvement of patients and their families in the development of oncology trials could encourage institutions to obtain these data in future protocols. Little published information exists on brain tumour presentation and diagnosis in adolescents and young adults. This is a population who have less parental supervision than younger children, in whom mood disturbance and behavioural change are common and in which individuals might be less willing to engage with health-care providers. Adolescents and young adults show a different tumour epidemiology to children and often have disturbances of growth and puberty.100 Therefore, we caution the application of information obtained from across all age groups to adolescents and young adults and, just as infants and young children are often reported separately from older children, they should be regarded as a separate age group. No published data exist on symptom progression in children with CNS tumours. If the diagnostic pathway is to be improved, more information is needed on the progression of symptoms and signs in children, from symptom onset to diagnosis and the factors that prompt imaging. Such studies would have to be retrospective, should obtain information directly from patients or carers, and will be relatively expensive; however, only by understanding how symptoms and signs of CNS tumours develop and the factors that lead to a request for imaging can more rapid diagnosis be achieved. This analysis shows both the heterogeneity of childhood CNS tumour presentation and the importance of tumour http://oncology.thelancet.com Vol 8 August 2007
location, age, and neurofibromatosis status in presentation. By ranking symptoms and signs and reporting by age and tumour location, this study focuses on the associative features in a hierarchical way that could be useful for training. Symptoms and signs of raised intracranial pressure occur in less than 50% of all children with intracranial tumours. Motor system abnormalities, especially abnormalities of gait and coordination, are common with all tumour types. Eye signs are common in all intracranial tumour types. Macrocephaly is common in children under 4 years who have intracranial tumours. Weight loss occurs with all tumour types, growth failure with central tumours, and precocious puberty in children with neurofibromatosis and intracranial tumours. Assessment of any child who presents with symptoms and signs that could result from a CNS tumour should include a thorough visual and motor system examination, assessment of growth (including head circumference in children under 4 years), and pubertal status. Specific multiple symptoms and signs (eg, in the combinations shown in figure 4), should alert the clinician to the possibility of a CNS tumour. Contributors DW, JC, CK, RG, and SW devised the study. JC provided statistical advice. SW undertook the data collection, data analysis, and drafted the manuscript. All authors commented on manuscript drafts and contributed to the final version. Conflicts of interest The authors declared no conflicts of interest. Acknowledgments SW was funded by a research grant from The Big Lottery Fund in conjunction with The Samantha Dickson Brain Tumour Trust. References 1 Office for National Statistics. Deaths by age, sex and underlying cause, 2003 registrations. Health Stat Q; quarterly 22. 2 UNICEF. Monitoring the situation of children and women. http://www.childinfo.org/ (accessed April 26, 2007). 3 Young G, Toretsky J, Campbell A, Eskenazi A. Recognition of common childhood malignancies. Am Fam Physician 2000; 61: 2144–54. 4 NICE. Referral guidelines for suspected cancer in adults and children. http://www.nice.org.uk (accessed April 26, 2007). 5 Cancer Research UK. News & resources. http://info. cancerresearchuk.org/cancerstats/ (accessed April 26, 2007). 6 Davison K, Ramsey M. The epidemiology of acute meningitis in children in England and Wales. Arch Dis Child 2003; 88: 662–64. 7 EURODIAB ACE Study Group. Variation and trends in incidence of childhood diabetes in Europe. Lancet 2000; 355: 873–76. 8 Aarsen F, Paquier P, Reddingius R, et al. Functional outcome after low-grade astrocytoma treatment in childhood. Cancer 2006; 106: 396–402. 9 Ilveskoski I, Pihko H, Wiklund T, et al. Neuropsychologic late effects in children with malignant brain tumors treated with surgery, radiotherapy and “8 in 1” chemotherapy. Neuropediatrics 1996; 27: 124–29. 10 Macedoni-Luksic M, Jereb B, Todorovski L. Long-term sequelae in children treated for brain tumors: impairments, disability, and handicap. Pediatr Hematol Oncol 2003; 20: 89–101. 11 Kennedy C, Leyland K. Comparison of screening instruments for disability and emotional/behavioral disorders with a generic measure of health-related quality of life in survivors of childhood brain tumors. Int J Cancer 1999; 12 (suppl): 106–11. 12 Martinez-Climent J, Sanchez V, Menor C, et al. Scale for assessing quality of life of children survivors of cranial posterior fossa tumors. J Neurooncol 1994; 22: 67–76.
693
Articles
13 14
15
16 17
18 19
20 21
22
23
24 25 26 27 28 29
30
31
32
33
34
35
36 37
38 39
40
694
Hansard. Brain cancer (children). London: House of Commons, 2006: column 825. Cancer: clinical features and management. Section 6.6 in: Warrell D, Cox T, Firth J, Benz E, eds. Oxford textbook of medicine, 4th edn. Oxford: Oxford University Press, 2004. Haslam RHA. Brain tumors in children. In: Nelson W, Behrman R, Kliegman R, Arvin A, eds. Textbook of pediatrics, 15th edn. Philadelphia: Harcourt Brace and Company, 1996: 1731–35. Dixon-Woods M, Findlay M, Young B, et al. Parents’ accounts of obtaining a diagnosis of childhood cancer. Lancet 2001; 357: 670–74. Childhood Brain Tumor Consortium. The epidemiology of headache among children with brain tumor. J Neurooncol 1991; 10: 31–46. Lipsey M, Wilson D. Practical meta-analysis. Thousand Oaks, CA, USA: SAGE publications, 2001. Czyzyk E, Jowiak S, Roszkowski M, Schwartz R. Optic pathway gliomas in children with and without neurofibromatosis 1. J Child Neurol 2003; 18: 471–78. Tewari M, Sharma B, Mahajan R, et al. Supratentorial tumours in infants. Child Nerv Syst 1994; 10: 172–75. Balestrini M, Micheli R, Giordano L, et al. Brain tumors with symptomatic onset in the first two years of life. Childs Nerv Syst 1994; 10: 104–10. Haddad S, Menezes A, Bell W, et al. Brain tumors occurring before 1 year of age: a retrospective reviews of 22 cases in an 11-year period (1977–1987). Neurosurgery 1991; 29: 8–13. Trujillo-Maldonado A, Davila-Gutierrez G, Escanero-Salazar A, et al. [Brain tumors in nursing infants]. Bol Med Hosp Infant Mex 1991; 48: 807–13. Lund A, Skovby F. Optic gliomas in children with neurofibromatosis type 1. Eur J Pediatr 1991; 150: 835–38. Haddad S, Moore S, Menezes A, VanGilder J. Ganglioglioma: 13 years of experience. Neurosurgery 1992; 31: 171–78. Smith N, Carli M, Hanieh A, et al. Gangliogliomas in childhood. Childs Nerv Syst 1992; 8: 258–62. Vandertop W, Hoffman H, Drake J, et al. Focal midbrain tumors in children. Neurosurgery 1992; 31: 186–94. Packer R, Parilongo G, Johnson D, et al. Choroid plexus carcinoma of childhood. Cancer 1991; 69: 580–85. Hope J, Armstrong D, Babyn P, et al. Primary meningeal tumors in children: correlation of clinical and CT findings with histologic type and prognosis. Am J Neuroradiol 1992; 13: 1353–64. Reed U, Rosemberg S, Gherpelli J, et al. Brain tumors in the first two years of life: a review of forty cases. Pediatr Neurosurg 1993; 19: 180–85. Hoppe-Hirsch E, Hirsch J. [Malignant tumors of the cerebral hemisphere in children. A series of 64 cases]. Arch Fr Pediatr 1993; 50: 403–07. Hung P, Wang H, Chou M, et al. Intracranial meningiomas in childhood. Zhonghua Min Guo Xiao Er Ke Yi Xue Hui 1994; 35: 495–501. Kuenzle C, Weissert M, Roulet E, et al. Follow-up of optic pathway gliomas in children with neurofibromatosis. Neuropediatrics 1994; 25: 295–300. Gordon G, Wallace S, Neal J. Intracranial tumours during the first two years of life: presenting features. Arch Dis Child 1995; 73: 345–47. Robertson P, Muraszko K, Brunberg J, et al. Pediatric midbrain tumors: a benign subgroup of brainstem gliomas. Pediatr Neurosurg 1995; 22: 65–73. Zhou B, Li J, Luo Z. [Clinicopathologic analysis of ependymoma in childhood]. Zhonghua Bing Li Xue Za Zhi 1995; 24: 369–71. Molloy PT, Bilaniuk LT, Vaughan SN, et al. Brainstem tumors in patients with neurofibromatosis type 1: a distinct clinical entity. Neurology 1995; 45: 1897–902. Edgeworth J, Bullock P, Bailey A, et al. Why are brain tumours still being missed? Arch Dis Child 1996; 74: 148–51. Kaplan A, Albright A, Zimmerman R, et al. Brainstem gliomas in children. A Children’s Cancer Group review of 119 cases. Pediatr Neurosurg 1996; 24: 185–92. Chintagumpala M, Armstrong D, Miki S, et al. Mixed neuronal-glial tumors (gangliogliomas) in children. Pediatr Neurosurg 1996; 24: 306–13.
41 42
43 44 45
46
47
48 49 50 51
52 53
54
55
56
57
58 59
60 61
62 63
64 65 66
67
68
Dirks P, Harris L, Hoffman H, et al. Supratentorial primative neuroectodermal tumours in children. J Neurooncol 1996; 29: 75–84. Constantino S, Houten J, Miller D, et al. Intramedullary spinal cord tumors in children under the age of 3 years. J Neurosurg 1996; 85: 1036–43. Parker A, Robinson R, Bullock P. Difficulties in diagnosing intrinsic spinal cord tumours. Arch Dis Child 1996; 75: 204–07. Bognar L. Brain tumours during the first year of life. Ann N Y Acad Sci 1997; 824: 148–55. Chitnavis B, Phipps K, Harkness W, Hayward R. Intrinsic brainstem tumours in childhood: a report of 35 children followed for a minimum of 5 years. Br J Neurosurg 1997; 11: 206–09. Behnke J, Christen H-J, Bruck W, Marakis E. Intra-axial endophytic tumors in the pons and/or medulla oblongata. I. Symptoms, neuroradiological findings, and histopathology in 30 children. Childs Nerv Syst 1997; 13: 122–34. Johnson J, Hariharan S, Berman J, et al. Clinical outcome of pediatric gangliogliomas: ninety-nine cases over 20 years. Pediatr Neurosurg 1997; 27: 203–07. Erdincler P, Lena G, Sarioglu A, et al. Intracranial meningiomas in children: review of 29 cases. Surg Neurol 1997; 49: 136–41. Lin I, Shu S, Chu H, Chi C. Primary intracranial germ-cell tumor in children. Chinese Med J 1997; 60: 259–64. Bilaniuk L, Molloy P, Zimmerman R, et al. Neurofibromatosis type 1: brain stem tumours. Neuroradiology 1997; 39: 642–53. Przybylski G, Albright A, Martinez A. Spinal cord astrocytomas: long-term results comparing treatments in children. Childs Nerv Syst 1997; 13: 375–82. Furuta T, Tabuchi A, Adachi Y, et al. Primary brain tumors in children under age 3 years. Brain Tumor Pathol 1998; 15: 7–12. Jovani Casano C, Canete Nieto A, Bermudez Cortes M, et al. [Central nervous system tumors in children less than three years of age]. An Esp Pediatr 1998; 49: 151–56. Chung S-K, Wang K-C, Nam D-H, Cho B-K. Brain tumor in the first year of life: a single institute study. J Korean Med Sci 1998; 13: 65–70. Young Poussaint T, Kowal J, Barnes P, et al. Tectal tumours of childhood: clinical and imaging follow-up. Am J Neuroradiol 1998; 19: 977–83. Berger C, Thiesse P, Lellouch-Tubian A, et al. Choroid plexus carcinomas in childhood: clinical features and prognostic factors. Neurosurgery 1998; 42: 470–75. Comi A, Backstrom J, Burger P, Duffner P. Clinical and neuroradiologic findings in infants with intracranial ependymomas. Pediatric Oncology Group. Pediatr Neurol 1998; 18: 23–29. Bouffet E, Pierre-Kahn A, Marchal J, et al. Prognostic factors in pediatric spinal cord astrocytoma. Cancer 1998; 83: 2391–99. Lonjon M, Goh K, Epstein F. Intramedullary spinal cord ependymomas in children: treatment, results and follow-up. Pediatr Neurosurg 1998; 29: 178–83. Keene D, Hsu E, Ventureyra E. Brain tumors in childhood and adolescence. Pediatr Neurol 1999; 20: 198–203. Sala F, Cloarusso E, Mazza C, et al. Brain tumors in children under 3 years of age. Recent experience (1987–1997) in 39 patients. Pediatr Neurosurg 1999; 31: 16–26. Kansal S, Jindal A, Mahapatra A. Brain stem glioma—a study of 111 patients. Indian J Cancer 1999; 36: 99–108. Grant G, Avellino A, Loeser J, et al. Management of intrinsic gliomas of the tectal plate in children. A ten-year review. Pediatr Neurosurg 1999; 31: 170–76. Zuccaro G, Sosa F, Cuccia V, et al. Lateral ventricle tumors in children: a series of 54 cases. Childs Nerv Syst 1999; 15: 774–85. Drummond K, Rosenfeld J. Pineal region tumours in childhood. A 30-year experience. Childs Nerv Syst 1999; 15: 119–26. Yang H, Nam D, Wang K, et al. Supratentorial primitive neuroectodermal tumor in children: clinical features, treatment outcome and prognostic factors. Childs Nerv Syst 1999; 15: 377–83. Young Poussaint T, Yousef N, Barnes P, et al. Cervicomedullary astrocytomas of childhood: clinical and imaging follow-up. Pediatr Radiol 1999; 29: 662–68. Thulesius H, Pola J, Hakansson A. Diagnostic delay in pediatric malignancies—a population-based study. Acta Oncol 2000; 39: 873–76.
http://oncology.thelancet.com Vol 8 August 2007
Articles
69
70
71
72 73
74
75
76
77
78
79
80
81 82 83
84 85
McEvoy A, Harding B, Phipps K, et al. Management of choroid plexus tumours in children: 20 years experience at a single neurosurgical centre. Pediatr Neurosurg 2000; 32: 192–99. Akyuz C, Emir S, Akalan N, et al. Intracranial ependymomas in childhood—a retrospective review of sixty-two children. Acta Oncol 2000; 39: 97–100. Amirjamshidi A, Mehrazin M, Abbassioun A. Meningiomas of the central nervous system occurring below the age of 17: report of 24 cases not associated with neurofibromatosis and review of literature. Childs Nerv Syst 2000; 16: 406–16. Hamelin N, Becquet P, Pierre-Kahn A, Dufier J-L. Clinical features of gliomas of the chiasma. J Fr Ophthalmol 2000; 23: 694–98. Mucic-Pucic B, Cvitanovic-Sojat L, Hajnzic T, Mataija M. Neurologic symptoms as first signs of brain tumour in children. Acta Clinica Croatica 2001; 40: 27–30. Due-Tonnessen B, Helseth E, Skullerud K, Lundar T. Choroid plexus tumors in children and young adults: report of 16 consecutive cases. Childs Nerv Syst 2001; 17: 252–56. Halperin E, Watson D, George S. Duration of symptoms prior to diagnosis is related inversely to presenting disease stage in children with medulloblastoma. Cancer 2001; 91: 1444–50. Nishio S, Morioka T, Taweshita I, Fukui M. Nerve cell tumours of the cerebrum: variable clinical and pathological manifestations. J Clin Neurosci 2001; 8: 225–30. Rohrer T, Gassman K, Buchfelder M, et al. [Clinical symptoms in 35 children and adolescents with craniopharyngeoma at the time of diagnosis]. Klin Padiatr 2002; 214: 285–90. Dobrovoljac M, Hengartner H, Boltshauser E, Grotzer MA. Delay in the diagnosis of paediatric brain tumours. Eur J Pediatr 2002; 161: 663–67. Mehta V, Chapman A, McNeely P, et al. Latency between symptom onset and diagnosis of pediatric brain tumors: an eastern Canadian geographic study. Neurosurgery 2002; 51: 365–72. Im S, Chung C, Cho B, et al. Intracranial ganglioglioma: preoperative characteristics and oncologic outcome after surgery. J Neurooncol 2002; 59: 173–83. Martinez-Lage J, Perez-Espejo M, Esteban J, Paza M. Thalamic tumors: clinical presentation. Childs Nerv Syst 2002; 18: 405–11. Zwerdling T, Dothage J. Meningiomas in children and adolescents. J Pediatr Hematol Oncol 2002; 24: 199–204. Rivera-Luna R, Medina-Sanson A, Leal-Leal C, et al. Brain tumors in children under 1 year of age: emphasis on the relationship of prognostic factors. Childs Nerv Syst 2003; 19: 311–14. Young H, Johnston H. Intracranial tumors in infants. J Child Neurol 2004; 19: 424–30. Dai A, Backstrom J, Burger P, Duffner P. Supratentorial primitive neuroectodermal tumors of infancy: clinical and radiologic findings. Pediatr Neurol 2003; 29: 430–34.
http://oncology.thelancet.com Vol 8 August 2007
Alston R, Newton R, Kelsey A, et al. Childhood medulloblastoma in northwest England 1954 to 1997: incidence and survival. Dev Med Child Neurol 2003; 45: 308–14. 87 Guillamo J, Creange A, Kalifa C, et al. Prognostic factors of central nervous system tumours in neurofibromatosis 1 (NF1): a retrospective study of 104 patients. Brain 2003; 126: 152–60. 88 King A, Listernick R, Charrow J, et al. Optic pathway gliomas in neurofibromatosis type 1: the effect of presenting symptoms on outcome. Am J Med Genet 2003; 122: 95–99. 89 Wanyoike P. Posterior cranial fossa tumours in children at Kenyatta National Hospital, Nairobi. East Afr Med J 2004; 81: 258–60. 90 Sharma M, Ralte A, Gaekwad S, et al. Subependymal giant cell astrocytoma: a clinicopathological study of 23 cases with special emphasis on proliferative markers and expression of p53 and retinoblastoma gene proteins. Pathology 2004; 36: 139–44. 91 Thiagalingam S, Flaherty M, Billson F, North K. Neurofibromatosis type 1 and optic pathway gliomas: follow-up of 54 patients. Ophthalmology 2004; 111: 568–77. 92 Perek D, Drogosiewicz M, Dembowska-Baginska B, et al. Diagnostic problems in children with primary brain tumours treated in Children’s Memorial Health Institute. Pediatria Polska 2005; 80: 29–36. 93 Pollack IF. Brain tumors in children. N Engl J Med 1994; 331: 1500–07. 94 Moher D, Cook D, Eastwood S, et al. Improving the quality of reports of meta-analyses of randomised controlled trials: the QUOROM statement. Quality of reporting of meta-analyses. Lancet 1999; 354: 1896–900. 95 Phillips B, Ball C, Sackett D, et al. Oxford Centre for Evidence-based Medicine levels of evidence. 2001. http://www.cebm.net/index. aspx?o=1025 (accessed April 26, 2007). 96 Linet M, Ries L, Smith M, et al. Cancer surveillance series: recent trends in childhood cancer incidence and mortality in the United States. J Natl Cancer Inst 1999; 91: 1051–58. 97 Rickert C, Werner, P. Epidemiology of central nervous system tumors in childhood and adolescence based on the new WHO classification. Childs Nerv Syst 2001; 17: 503–11. 98 Wilne S, Ferris R, Nathwani A, Kennedy C. The presenting features of brain tumours: a review of 200 cases. Arch Dis Child 2006; 91: 502–06. 99 Davies E, Clarke C. Early symptoms of brain tumours. J Neurol Neurosurg Psychiatry 2004; 57: 1205–06. 100 Levy A. Brain tumors in children: evaluation and management. Curr Prob Pediatr Adolesc Health Care 2005; 35: 230–45. 86
695
Presentation of childhood CNS tumours: a systematic review and meta-analysis Sophie Wilne, Jacqueline Collier, Colin Kennedy, Karin Koller, Richard Grundy, David Walker
Summary Background Suspicion of a CNS tumour is classically raised by symptoms of raised intracranial pressure, focal deficits (including seizures), or papilloedema. Development of guidelines is needed for the identification and referral of children who might have a CNS tumour. We did a systematic literature review and meta-analysis to identify the clinical presentation of childhood CNS tumours to provide evidence to support the development of guidelines to assist with the identification and referral for imaging of children who might have a central nervous system tumour. Methods Medline, Embase, and PubMed were searched for cohort studies and case series in children, published between January, 1991, and August, 2005, detailing the symptoms and signs at diagnosis of a CNS tumour. Findings 74 papers (n=4171) met the inclusion criteria. 56 symptoms and signs at diagnosis were identified, ranked by frequency, and clustered according to age, anatomical criteria, and genetic criteria. The most frequent symptoms and signs at diagnosis were: headache (33%), nausea and vomiting (32%), abnormalities of gait and coordination (27%), and papilloedema (13%) for intracranial tumours; macrocephaly (41%), nausea and vomiting (30%), irritability (24%), and lethargy (21%) for children aged under 4 years with intracranial tumours; reduced visual acuity (41%), exophthalmia (16%), and optic atrophy (15%) for children with an intracranial tumour and neurofibromatosis; nausea and vomiting (75%), headache (67%), abnormal gait and coordination (60%), and papilloedema (34%) for posterior fossa tumours; unspecified symptoms and signs of raised intracranial pressure (47%), seizures (38%), and papilloedema (21%) for supratentorial tumours; headache (49%), abnormal eye movements (21%), squint (21%), and nausea and vomiting (19%) for central brain tumours; abnormal gait and coordination (78%), cranial nerve palsies (52%), pyramidal signs (33%), headache (23%), and squint (19%) for brainstem tumours; and back pain (67%), abnormalities of gait and coordination (42%), spinal deformity (39%), focal weakness (21%), and sphincter disturbance (20%) for spinal-cord tumours. Other features noted were weight loss, growth failure, and precocious puberty. Symptoms of raised intracranial pressure were absent in more than half of children with brain tumours. Other neurological features were heterogeneous and related to tumour location. Interpretation Apart from raised intracranial pressure, motor and visual system abnormalities, weight loss, macrocephaly, growth failure, and precocious puberty also suggest presence of an intracranial tumour. Children with signs and symptoms that could result from a CNS tumour need a thorough visual and motor system examination and an assessment of growth and pubertal status. Occurrence of multiple symptoms and signs should alert clinicians to possible CNS tumours.
Introduction Life-threatening clinical conditions in childhood are seen infrequently in developed countries.1,2 Screening of the few serious diagnoses from the many self-limiting conditions and fluctuations in developmental processes and behaviour is a major diagnostic challenge in primary and secondary health care.3,4 Cancer affects one in 600 children under 16 years and thus represents a moderate health risk similar to cerebral palsy, diabetes mellitus, and meningitis.5–7 A quarter of childhood cancers arise in the CNS and account for the largest number of cancer deaths in childhood.1 60% of survivors are left with pronounced disability.8–12 CNS tumours are therefore common in the context of life-threatening childhood disease. All healthcare professionals seeing children should be able to identify symptoms and signs that could result from a CNS tumour and refer or investigate appropriately. Despite advances in neuroimaging, the timely diagnosis of CNS tumours in the UK remains difficult.13 http://oncology.thelancet.com Vol 8 August 2007
Lancet Oncol 2007; 8: 685–95 Published Online July 18, 2007 DOI:10.1016/S14702045(07)70207-3 See Reflection and Reaction page 664 Children’s Brain Tumour Research Centre, Academic Division of Child Health (S Wilne MBBS, K Koller PhD, Prof R Grundy MBChB, Prof D Walker MBBS), and School of Nursing, Faculty of Medicine & Health Sciences (Prof J Collier PhD), University of Nottingham, Queen’s Medical Centre, Nottingham; and Department of Child Health, University of Southampton, Southampton General Hospital, Southampton, UK (Prof C Kennedy MBBS) Correspondence to: Dr Sophie Wilne, Children’s Brain Tumour Research Centre, Academic Division of Child Health, University of Nottingham, Queens Medical Centre, Nottingham NG7 2UH, UK sophie.wilne@nottingham. ac.uk
The varied presentation and perceived rarity (resulting in a low priority in the differential diagnosis) of CNS tumours in childhood underlie this problem. Many of the initial symptoms and signs of CNS tumours also occur with other more common and less serious childhood disorders such as gastroenteritis, migraine, and behavioural problems. Conventional teaching is that CNS tumours present with symptoms of raised intracranial pressure (early morning headache with vomiting and papilloedema) with or without focal neurological signs.14,15 Research in the pre-CT era has identified comprehensive lists of symptom and sign clusters with which neurologists and other paediatric specialists are highly familiar. Currently, expanded access to neuroimaging and the pressure to accelerate cancer diagnoses is placing a much broader group of doctors in the position to initiate imaging either directly or through an urgent cancer referral. A period of diagnostic uncertainty often precedes the diagnosis of a 685
Articles
Data collection MEDLINE, PubMed, and EMBASE searched from January, 1991, to August, 2005 5620 potentially relevant studies identified and abstracts screened for retrieval
5234 abstracts excluded because subject not related to paediatric CNS tumour presentation
386 papers retrieved for detailed review
312 papers excluded 32 had combined adult and paediatric data 4 had no paediatric data 9 had less than ten children 20 had insufficient detail on signs and symptoms 7 had combined CNS and extradural tumours 4 had no primary data 236 had unrelated subject matter or duplicate study
74 papers included in meta-analysis
Figure 1: Progress through the meta-analysis
CNS tumour, which patients and their families find extremely distressing, and any perception that the medical response has been inadequate, incompetent, or delayed may be associated with legal dispute.16 In 1991, the Childhood Brain Tumour Consortium17 reported the symptoms and signs at diagnosis for 3291 children diagnosed with a brain tumour in 1930–79. To provide a more contemporary cohort, we undertook a systematic review and meta-analysis of the presenting symptoms and signs in paediatric CNS tumours detailed in subsequent publications. The literature review was done as the initial stage in a project devising guidance to help health-care professionals identify children who might have a CNS tumour and who thus need fast-track imaging.
Methods Identification of studies and inclusion criteria We searched MEDLINE, PubMed, and EMBASE without language restriction, from January, 1991, to August, 2005. Key words were: “brain tumour(s), “brain tumor(s)”, “brain neoplasm(s)”, “spinal cord tumour(s)”, “spinal cord tumor(s)”, “spinal cord neoplasm”; and “diagnosis”; and “sign(s)” or “symptom(s)”. Retrieved references were restricted to “all child”. Abstracts were screened by a researcher (SW); those unrelated to CNS tumours or discussing an area unrelated to clinical presentation were excluded. Papers with abstracts discussing tumour presentation, tumour diagnosis, or clinical symptoms and signs were retrieved for detailed review. We included all case-series or cohort studies describing symptoms and signs at diagnosis for a minimum of ten children diagnosed with a CNS tumour and published after February, 1991. Non-English language papers were translated. 686
Numbers of children in every study with a symptom or sign at diagnosis were recorded on a standard data extraction form. Information on symptoms and signs varied between studies. Some studies had very detailed records on individual symptoms and signs (eg, headache, vomiting, papilloedema), whereas others reported symptoms in clusters or complexes (eg, symptoms of raised intracranial pressure). Symptoms and signs were recorded as described in the individual studies. If a symptom or sign was not recorded in a study, we assumed it not to occur in that population.
Statistical analysis Analysis was done with meta-disc version β 1.1.1. We combined proportions (%) of children with each symptom or sign at diagnosis using one-variable relationship metaanalysis. The effect size for each symptom and sign was calculated in the individual studies and weighted according to its variance, and these effect sizes were then summed (for each symptom and sign) and the total effect size was then divided by the sum of the weights to give a mean effect size (pooled proportion).18 In meta-disc, proportions (as well as likelihood ratios and diagnostic ratios) could be pooled with either the Mantel-Haenszel method (fixed-effects model) or, to incorporate variation between studies, with the DerSimonian Laird method (random-effects model). In the present study, heterogeneity was indicated beyond what could be expected by chance alone, by significant Q statistics and high inconsistency (I²) statistics. The DerSimonian Laird method was selected because variability was expected across the papers, and a randomeffects model was used.18 Symptoms and signs occurring in 5% or more of the meta-analysis population were reported. Two papers19,20 reported optic atrophy and papilloedema and one paper21 lethargy and irritability as a combined category. Since these papers reported detailed information for other symptoms and signs, they were included in the meta-analysis but excluded from the analysis of the combined symptoms or signs. In one report,19 visual acuity was not assessed in the complete cohort and, therefore, was excluded from the metaanalysis of visual acuity. The following subgroup analyses were undertaken: all intracranial tumours; intracranial tumours in children aged under 4 years; children with an intracranial tumour and neurofibromatosis; posterior fossa tumours; supratentorial (excluding central) tumours; central tumours (third ventricle, tectum, pineal gland, pituitary gland, thalamus, hypothalamus, optic pathway, and basal ganglia); brainstem tumours; and spinal-cord tumours. Analysis of all intracranial tumours was undertaken to provide a summary of paediatric intracranial tumour presentation. Children aged under 4 years usually cannot clearly describe symptoms such as headache, nausea, and diplopia, and therefore have a different presentation to http://oncology.thelancet.com Vol 8 August 2007
Articles
Recruitment Number of period patients
Patient group, diagnosis if known, source of data
Tumour location
Mean age, years
Median age, years
Age range, years
Mean symptom Median symptom Symptom Ref interval, interval, interval range, months months months NS
1977–87
22
Infants, 1I*
AB
NS
NS
NS
1981–89
16
Under 2 years, 1I*
AB
NS
NS
NS
NS
1965–89
16
NF1 and OP tumours, 2I
OP
NS
4·5
1·5–17
1978–91
12
Gangliogliomas, 1I
AB
NS
1977–90
12
Gangliogliomas, 1I
AB
1976–91
12
Midbrain tumours, 1I
C
1975–81
11
Choroid plexus carcinoma, 2I
AB
1976–88
21
Meningeal tumours, 1I
AB
NS
0·3–16·7
14·6
1962–89
39
Under 2 years, 1I*
AB
NS
NS
NS
NS
NS
NS
30
1970–89
106
Cerebral hemisphere tumours, 1I
ST
NS
NS
NS
NS
NS
NS
31
1970–87
80
Under 2 years at symptom onset, 1I*
AB
NS
1989–92
14
Infants with supratentorial tumours, 1I
ST
0·5
1980–90
10
Meningiomas, 1I
AB
1973–92
21
NF1 and OP tumours, 4I
OP
1979–94
21
Under 2 years, 1I*
AB
1983–92
17
Midbrain tumours, 1I
C
1974–94
23
Intracranial ependymoma, 1I
AB
1984–94
17
NF1 and BS tumours, 1I
BS
1990–94
74
AB tumours, 1I
AB
1988–91
119
BS gliomas treated with hyperfractionated radiotherapy (Children’s Cancer Group trial 9882)
BS
2·5
1
25·2
NS
22
0·5–9
23
NS
NS
24
NS
3·5–17
NS
NS
NS
25
7·8
NS
0·8–15·8
40
27
7–96
26
8·2
NS
1·1–16
2·2
0–9·5
NS 9·3
4·5 NS
NS
0·5–14
27
NS
NS
28
0–72
29
4
NS
NS
NS
NS
0–153·6
21
NS
0·1–0·9
NS
NS
NS
20
11·1
NS
8–15
13·2
NS
0·1–60
32
7·1
NS
0–14·5
NS
NS
NS
33
NS
NS
0·2–1·8
NS
NS
NS
34
NS
9·7
3·5–16
NS
4
NS
35
8·8
NS
2–14
8·4
8·3
1·3–13·9
6·9
NS
NS
6·5
NS
NS
NS
3·8 NS 4·6
NS
0·5–10
36
NS
NS
37
NS
0·2–30
38
NS
NS
39
1984–93
32
Gangliogliomas, 1I
AB
6·5
NS
0·7–20
NS
NS
NS
40
1970–95
36
Supratentorial primitive neuroectodermal tumours, 1I
ST
4·3
2·9
0·1–12·8
NS
NS
NS
41
1980–93
27
Under 3 years with intramedullary SC, 1I
SC
1·7
NS
0·5–3
NS
NS
NS
42
1984–95
13
Intrinsic SC tumours, 1I
SC
5·4
NS
0·7–11
NS
NS
NS
43
1984–95
723
AB tumours, 1I
AB
NS
NS
0–16
NS
NS
NS
44
1980–90
35
BS tumours, 1I
BS
NS
NS
1·3–13
5
NS
NS
45
1987–94
30
Endophytic pons or medullary tumours, 1I
BS
NS
6
0·6–16
6
NS
1–60
46
1974–95
99
Gangliogliomas, 1I
AB
NS
1·7–20
60
24·4
NS
47
1968–94
29
Meningiomas 2I
AB
10
NS
0–15
NS
NS
NS
48
1983–95
12
Primary intracranial germ-cell tumours, 1I
C
NS
NS
5–15
NS
NS
NS
49
1984–96
25
NF1 and BS tumours, 1I
BS
7·8
NS
1·1–15·2
NS
NS
NS
50
1976–92
18
SP astrocytomas, 1I
SC
9·2
8·6
0·6–17·9
NS
NS
NS
51
1966–96
46
Under 3 years, 1I*
AB
NS
NS
NS
NS
NS
52
1985–95
20
Under 3 years, 1I*
AB
1·7
NS
0–2·7
1977–96
21
Infants, 1I
AB
0·5
NS
NS
NS
1990–97
32
Tectal tumours, 1I
C
8
NS
0·2–17
1984–95
22
Choroid plexus carcinoma registered with Societe Francais Oncologie Paediatric
AB
NS
2·1
0·3–9·3
1986–90
40
Intracranial ependymoma treated in Paediatric Oncology Group trial 8633
AB
NS
NS
0·3–2·9
1971–94
73
SP astrocytomas, 13I
SC
NS
7
0·3–6
NS
NS
NS
58
1985–96
20
Intramedullary SP ependymomas, 1I
SC
14
NS
9–18
NS
NS
NS
59
1975–93
200
All brain tumours, 1I
AB
8·9
NS
NS
NS
NS
NS
60
1987–97
39
Under 3 years, 1I*
AB
2·1
NS
0·3–3
NS
0·2–18
61
1983–97
76
BS gliomas
BS
NS
NS
3–15
NS
NS
NS
62 63
9·5
NS
2·4
0–18
53
NS
NS
54
NS
NS
NS
55
NS
1
0·1–8
56
1
0–10·9
57
1·6
5·2
1
1988–98
11
Tectal-plate gliomas, 1I
C
10
NS
5–13
28·2
NS
0·7–84
1988–98
54
Lateral-ventricle tumours, 1I
ST
NS
NS
0–20
5
NS
0–48
64
1967–97
37
Pineal-region tumours, 1I
C
9·6
NS
NS
NS
NS
NS
65
1986–95
28
Supratentorial primitive neuroectodermal tumours, 1I
ST
6·8
NS
0·7–16·9
1988–98
11
Cervicomedullary astrocytomas, 1I
SC
7
NS
0–18
4·9 NS
NS
1–48
66
NS
NS
67
(Continues on next page)
http://oncology.thelancet.com Vol 8 August 2007
687
Articles
(Continued from previous page) 1984–95
22
Any brain tumours reported to regional tumour registry
AB
NS
1979–99
34
Choroid plexus tumours, 1I
AB
NS
1·4
0·1–11·5
NS
1
1972–91
62
Intracranial ependymoma, 1I
PF
6
NS
1–17
2
NS
1984–99
24
Meningiomas, 2I
AB
NS
NS
2–17
8·2
NS
0·2–14·4
71
1980–94
18
Chiasmal gliomas, 1I
OP
NS
NS
0·5–14
NS
NS
NS
72
1985–99
181
AB tumours, 1I
AB
NS
NS
0–16
NS
NS
NS
73
1970–98
16
Choroid plexus tumours, 1I
AB
NS
0·2–15·4
NS
NS
NS
74
1974–99
122
Medulloblastoma, 1I
PF
NS
NS
NS
NS
NS
75
1981–98
11
Nerve-cell tumours, 1I
ST
NS
NS
2–16
NS
NS
NS
76
1970–98
35
Craniopharyngiomas, 1I
C
NS
9·1
1·3–15·6
NS
NS
NS
77
1980–99
252
AB, 1I
AB
NS
NS
1995–2000
104
AB, 2I
AB
1987–99
22
Gangliogliomas, 2I
AB
1980–2000
20
Thalamic and basal ganglia tumours, 1I
C
1974–99
18
Meningiomas recorded in a hospital tumour registry
AB
1975–2002
61
Infants, 2I*
AB
1988–99
16
Infants, 1I*
AB
1986–90
13
Supratentorial primative neuroectodermal tumours treated in Pediatric Oncology Group trial 8633
1954–97
181
1982–2000 1996–2000
69
3·1
8·29 NS 6·6
NS
6·3
0–16·9
NS
NS
NS
0–16
NS
0·3–18
4·6
3·3
7·3 30 1·5
0·2–45·9
68
0·03–33
69
NS
70
1·8
0–99
78
3
NS
79
11
NS
80
NS
0–24
81
NS
NS
82
NS
0·1–8·9
83
0.2
0–6
84
NS
0–49
85
1·6–17
NS
0–1
NS
0·5
0–1
ST
NS
NS
0–3
Medulloblastoma registered with Manchester Children’s tumour registry
PF
NS
NS
0–14
NS
NS
NS
86
NF1 and symptomatic tumours, 7I
AB
NS
5·2
0·3–17
NS
NS
NS
87
OP
NS
NS
0·3–17·4
NS
NS
NS
19
NS
NS
NS
88
NS
NS
89
2·5–96
90
0·5
1986–2002
51
NF1 and symptomatic optic pathway gliomas, 2I
OP
4·8
NS
0–15·8
1996–2003
37
PF tumours, 1I
PF
6·7
NS
2–16
1978–2001
18
Giant-cell astrocytomas, 2I
AB
1973–2002
57
NF1 and OP tumours, 1I
OP
1997–2000
172
AB, 1I
AB
NS 5·2 NS
NS
2·1
NS
83 (51 NF1) OP gliomas, 2I
11
NS
1·9 NS 0·9
3·7
NS
4–15
19
9
NS
NS
NS
NS
8·3
0·3–17·3
4·9
1
NS
91
0·2–120
92
1I=treated at one institution. 2I=treated at two institutions. 4I=treated at four institutions. 7I=treated at seven institutions. AB=all brain. NS=not specified. OP=optic pathway. C=central. ST=supratentorial. BS=brainstem. SC=spinal cord. PF=posterior fossa. *Study population defined by age rather than tumour type or location.
Table: Studies meeting inclusion criteria
older children. Neurofibromatosis is the commonest genetic abnormality associated with intracranial tumours and children can develop tumours before the development of cutaneous manifestations. Children with neurofibromatosis have a high occurrence of optic-pathway tumours, and thus their presentation differs from that of other children with intracranial tumours. Only children with neurofibromatosis and a symptomatic intracranial tumour were included in this subgroup analysis. We did not analyse asymptomatic children with an intracranial tumour identified by CNS imaging that was instigated after a diagnosis of neurofibromatosis. Analysis by tumour location was undertaken to highlight specific associations of symptoms and signs that occur with different tumour locations.
Role of the funding source The sponsors of the study had no role in study design, data collection, data analysis, or data interpretation, in the writing of this report, or in the decision to submit the paper for publication. All authors had full access to all 688
the data in the study, and the corresponding author had final responsibility for the decision to submit for publication.
Results 5620 papers were identified by the search terms. We reviewed 386 papers in full, from which 74 met inclusion criteria, describing the symptoms and signs at diagnosis in 4171 children (figure 1, table).19–92 56 symptoms and signs were recorded in children with CNS tumours, but only symptoms and signs that occurred in 5% or more of patients are reported here. 61 studies (n=3702)19–23,25–32,34–36,38–41,44–49,52–57,60–66,68–86,89,90,92 described the symptoms and signs at diagnosis for children without neurofibromatosis who had an intracranial tumour. These were (in decreasing order of frequency): headache (33%), nausea and vomiting (32%), abnormal gait or coordination (27%), papilloedema (13%), seizures (13%), unspecified symptoms and signs of raised intracranial pressure (10%), squint (7%), change in behavioural or school performance (7%), macrocephaly (7%), cranial http://oncology.thelancet.com Vol 8 August 2007
Articles
nerve palsies (unspecified; 7%), lethargy (6%), abnormal eye movements (nystagmus, Parinaud’s syndrome; 6%), hemiplegia (6%), weight loss (5%), focal motor weakness (5%), unspecified visual or eye abnormalities (5%), and altered level of consciousness (5%; figure 2). 13 studies (n=332)20–23,30,34,52–54,57,61,83,84 were included in the analysis of children with intracranial tumours aged under 4 years. Ranked symptoms and signs at diagnosis were: macrocephaly (41%), nausea and vomiting (30%), irritability (24%), lethargy (21%), abnormal gait and coordination difficulties (19%), weight loss (14%), clinically apparent hydrocephalus (bulging fontanelle, splayed sutures; 13%), seizures (10%), papilloedema (10%), headache (10%), unspecified focal neurological signs (10%), unspecified symptoms of raised intracranial pressure (9%), focal motor weakness (7%), head tilt (7%), altered level of consciousness (7%), squint (6%), abnormal eye movements (6%), developmental delay (5%), and hemiplegia (5%; figure 2) Eight studies (n=307)19,24,33,37,50,87,89,92 were included in the analysis of children with neurofibromatosis and an intracranial tumour. The most common symptom and signs at diagnosis were visual, indicating the high occurrence of optic pathway gliomas in this population. The ranked symptoms and signs were reduced visual acuity (41%), exophthalmia (16%), optic atrophy (15%), squint (13%), headache (9%), unspecified symptoms of raised intracranial pressure (8%), precocious puberty (8%), abnormal gait or coordination difficulties (7%), voice abnormalities (6%), developmental delay (5%), papilloedema (5%), and reduced visual fields (5%; figure 2). Five studies (n=476)60,70,75,86,89 described children with posterior fossa tumours; seven studies (n=303)20,31,41,60,64,66,85 described children with supratentorial tumours; 11 (n=276)19,27,35,49,55,60,63,65,72,77,81 described children with central tumours; five (n=276)39,45,46,60,62 described children with brainstem tumours; and six studies (n=162)42,51,58,59,67,74 described children with spinal-cord tumours (figures 3 and 4).
Discussion To our knowledge, our study is the first systematic review and meta-analysis of symptom and sign patterns in children with CNS tumours. Our findings show the importance of age, tumour location, and neurofibromatosis status in the determination of the symptom and sign clusters present at tumour diagnosis. Combining the most common specific symptom or sign of raised intracranial pressure with the proportion of children presenting with non-specific symptoms or signs of raised intracranial pressure provides an estimate of the overall frequency of these symptoms and signs. This relation indicates that symptoms linked to raised intracranial pressure are present in about 40% of all intracranial tumours, 40% of intracranial tumours in children aged under 4 years, 20% of intracranial tumours occurring in http://oncology.thelancet.com Vol 8 August 2007
All intracranial tumours
Children under 4 years
Children with neurofibromatosis
Headache Nausea and vomiting Papilloedema Macrocephaly Clinically apparent hydrocephalus Symptoms of raised ICP NOS Abnormal gait or coordination Hemiplegia Focal motor weakness Head tilt Squint Abnormal eye movements Reduced visual acuity Reduced visual fields Optic atrophy Exophthalmia Visual or eye signs NOS Impaired consciousness Lethargy Behavioural or school change Irritability Weight loss Developmental delay Voice abnormalities CNP NOS Seizures Precocious puberty Focal neurological signs NOS Pooled proportion 95% CI
0
10 20 30 40 50 0
10 20 30 40 50 0
10 20 30 40 50
Pooled proportion of children affected (%)
Figure 2: Frequency of symptoms and signs in children with intracranial tumours Analysis by age and neurofibromatosis status. ICP=intracranial pressure. NOS=not otherwise specified. CNP=cranial nerve palsy.
children with neurofibromatosis, 80% of posterior fossa tumours, 60% of central tumours, 60% of hemispheric tumours, 30% of brainstem tumours, and 7% of spinalcord tumours. Other alerts to a possible CNS tumour include abnormal gait and coordination, other motor system abnormalities, eye signs, weight loss, behavioural changes (including lethargy and irritability) and school difficulties, developmental delay, cranial nerve palsies, head tilt, macrocephaly, diabetes insipidus, and growth arrest. Increasing awareness of the varied and complex symptomatology that often occurs with CNS tumours could help tumour diagnosis and reduce the extended symptom interval experienced by many children. Recognition that specific combinations of symptoms and signs indicate a focal CNS lesion is crucial to the diagnosis of many CNS tumours. 45–60% of childhood brain tumours are infratentorial, 25–40% are hemispheric, and 15–20% are midline supratentorial.93 Our approach has emphasised the symptom and sign combinations that occur with different tumour locations and will help 689
690
95% CI
Pooled proportion
Sphincter disturbance
Spinal deformity
Back pain
Accidental head injury
Focal neurological signs NOS
Short stature
Diabetes insipidus
Seizures
Cranial nerve palsies
Auditory symptoms or vertigo
Facial palsy
Developmental delay
Weight loss
Behavioural or school change
Lethargy
Altered level of conciousness
Visual or eye signs NOS
Optic atrophy
Reduced visual fields
Reduced visual acuity
Abnormal eye movements
Squint
Stiff neck
Decreased upper limb movement
Head tilt
Focal motor weakness
Pyramidal signs
Hemiplegia
Abnormal gait or coordination
Symptoms of raised ICP (NOS)
Macrocephaly
Papilloedema
Nausea without vomiting
Nausea and vomiting
Headache
0
10 20 30 40
50 60 70 80 90 100 0
Posterior fossa tumours
10 20 30 40
50 60 70 80 90 100 0
Supratentorial (excluding central) tumours
50 60 70 80 90 100 0 Pooled proportion of children affected (%)
10 20 30 40
Central tumours
10 20 30 40
50 60 70 80 90 100 0
Brain-stem tumours
10 20 30 40
50 60 70 80 90 100
Spinal-cord tumours
Articles
Figure 3: Frequency of symptoms and signs in children with CNS tumours Analysis by tumour location. ICP=intracranial pressure. NOS=not otherwise specified.
http://oncology.thelancet.com Vol 8 August 2007
Articles
Brain stem tumours:
Supratentorial tumours: Unspecified symptoms of raised ICP* 47% Seizures 38% Papilloedema* 21% Focal neurological signs 17% Headache* 11% Hemiplegia 10% Nausea and vomiting* 8% Macrocephaly* 6%
Abnormal gait and coordination difficulties 78% Cranial nerve palsies (unspecified) 52% Pyramidal signs (unspecified) 33% Headache* 23% Squint 19% Focal motor weakness 19% Facial palsy 15% Papilloedema* 13% Unspecified symptoms of raised ICP* 10% Abnormal eye movements 6% Behavioural change or school difficulties 5%
Posterior fossa tumours: Nausea and vomiting* 75% Headache* 67% Abnormal gait and coordination difficulties 60% Papilloedema* 34% Abnormal eye movements 20% Lethargy 13% Nausea without vomiting* 10% Unspecified symptoms and signs of raised ICP* 9% Weight loss 9% Focal motor weakness 9% Macrocephaly* 7% Impaired consciousness 7% Vertigo or auditory symptoms 7% Squint 6% Stiff neck 6% Head tilt Accidental head injury 5%
Spinal cord tumours: Back pain 67% Abnormal gait or coordination difficulties 42% Spinal deformity 39% Focal motor weakness 21% Sphincter disturbance 20% Decreased upper limb movement 17% Developmental delay 8% Head tilt 7% Headache* 7%
Central tumours: Headache* 49% Abnormal eye movements and squint 21% Nausea and vomiting* 19% Papilloedema* 18% Reduced visual acuity 16% Unspecified symptoms and signs of raised ICP* 13% Diabetes insipidus 12% Abnormal gait and coordination difficulties 10% Optic atrophy 9% Behavioural change or school difficulties 9% Altered level of consciousness 9% Reduced visual fields 8% Seizures 7% Hemiplegia 7% Focal motor deficit 7% Developmental delay 7% Short stature 7% Weight loss 5% Vertigo or auditory symptoms 5% Visual or eye abnormalities (unspecified) 5%
Figure 4: CNS tumour presentation *Symptom or sign caused by raised intracranial pressure (ICP).
focus the search for corroborative findings in children who present with a symptom or sign that is potentially suggestive of a CNS tumour. In many instances, the possibility that the symptoms or signs are the result of a CNS tumour will be (rightly) rapidly dismissed. However, consideration of this diagnosis in some cases could lead to identification of corroborative symptoms and signs and the instigation of imaging. Even if an underlying tumour is unlikely, patients and their families or carers should be encouraged to return for re-assessment should symptoms or signs persist or progress, and the diagnosis should be reviewed on re-presentation. We chose a 5% threshold for reporting symptoms and signs in children with CNS tumours as a practical compromise between the need to consider an underlying CNS tumour with a clinical feature not associated with this tumour type and those symptoms and signs that occur frequently in childhood CNS tumours. Because of the differing presentation of CNS tumours according to patient group and tumour location, most symptoms and http://oncology.thelancet.com Vol 8 August 2007
signs that occurred in less than 5% of patients in one subgroup occurred more frequently in the other subgroups. Symptoms and signs that consistently occurred in less than 5% of patients, which could be associated with diagnostic difficulty, were dysphagia and delayed puberty. 35 studies21,23,24,26,27,29,32,35,36,38,45-47,53,56,57,61,63,64,66,68–71,75,78–81,83–85,89,90,92 meeting the inclusion criteria reported symptom interval duration (table). Symptom interval comparison is difficult for several reasons. Studies report different measures of symptom interval (median, mean, range) and rarely report all three, and the statistical significance of any differences in symptom interval cannot be determined from the reported data. For asymmetric distributions such as symptom interval of paediatric brain tumours, the median provides the best comparator. The reported median symptom interval ranges from 1 to 27 months. The longest median symptom interval occurs with biologically slowgrowing tumours such as gangliogliomas,26,47 although there is little association between symptom interval range 691
Articles
and tumour biology, indicating that extended symptom intervals could occur with all types of paediatric brain tumour. Any association between specific symptoms and signs and an extended symptom interval could not be identified in this study; exploration of any such relation would need a large cohort study in which the symptom interval and time of onset for individual symptoms and signs was recorded for all patients. We used a systematic search strategy and standardised inclusion criteria, as recommended in the quality of reporting meta-analyses (QUOROM) statement, to identify studies for inclusion.94 The high number of papers identified in the past 15 years shows the sustained interest in the mechanisms of diagnosis in this group of patients. Previously published evidence on paediatric CNS tumour presentation has been predominantly in the form of case studies (level 4 evidence) with infrequent tumour registry series (level 2 evidence; two studies met the inclusion criteria for this study).68,86,95 The systematic approach has generated a cohort of patients, most of whom were diagnosed during the era of CT and MRI, six times larger that the largest single identified study. The meta-analysis results reported here provide level 2 evidence for this cohort, which give greater value to the rankings of symptoms and signs by age, tumour location, and neurofibromatosis status than previous reports. This study had some important limitations and potential sources of bias. The search strategy might not have identified all relevant papers and unpublished data were not sought. Papers included in the analysis reported symptoms and signs at diagnosis in children with a CNS tumour; therefore accuracy of these data depends on the history given by patients and their families or carers and the signs detected by the examining health-care practitioners. However, medical decisions will always be based on such histories and examination findings rather than the underlying full facts to which they relate. The assumption was made that if a symptom or sign was not described in a study, it did not occur in that population. The variability and large number of included patients should reduce the risk that common symptoms and signs are underrepresented and uncommon ones over-represented. We saw variation in the data detail between studies. Some studies were very detailed, recording individual symptoms and signs such as headache, vomiting, and papilloedema; 27–29,32,34–36,38,40,44,47,53,55,57,60–63,65,66,68,70,73,75–82,85,86,89,92 whereas others used symptom complexes such as symptoms of raised intracranial pressure or cranial nerve palsies.19,31,39,41,48,52,56,60,64,69,71,83 Some symptoms and signs could have been combined to indicate the total proportion of children presenting with a specific symptom complex. However, since we could not clearly deduce exactly how the data related, some inaccuracy and misrepresentation of data could result and thus we kept data in their original form. Despite these problems, the analysis shows the variability of symptoms and signs and the frequency with which they occur in childhood CNS tumours. 692
Most childhood brain tumours are low-grade astrocytomas.96,97 Apart from optic pathway gliomas, these astrocytomas were under-represented in the studies identified. This result is probably due to a historical failure to include non-malignant brain tumours in tumour registries and, until recently, absence of review of children with low-grade gliomas by paediatric oncologists. Despite this result, the distribution of tumour location in the studies identified here was similar to that seen in clinical practice (56% infratentorial, 23% hemispheric, 21% central), lending support to the analysis results. Publication bias could have led to overrepresentation of rare tumours or those with an unusual presentation; however, we excluded case reports and studies with fewer than ten patients to combat this problem. Finally, this analysis addresses the issue of sensitivity but not that of specificity of symptoms and signs to the presence of an underlying CNS tumour. Therefore, our findings might be more useful in guiding clinical-management decisions in children who are already selected by referral to secondary or tertiary care than in unselected children. The probability of a symptom or sign being indicative of a CNS tumour will increase with the occurrence of corroborative findings on history and examination and the prevalence of CNS tumours in the population in question. The previous largest study17 of childhood brain tumour presentation, undertaken by the Childhood Brain Tumor Consortium, reported on the distribution of other symptoms and neurological signs in 3291 children with or without headache in association with a brain tumour. For the most of this period, CT and MRI were not available. Direct comparison between this study17 and the present analysis is complicated by differences in anatomical subdivision and methodology. Headache, nausea and vomiting, and seizures were reported for the entire cohort, although other symptoms were reported for specific age groups and numbers in each age group were not provided.17 Similarly, although the occurrence of coma, focal motor weakness, and papilloedema is reported for the entire group, other symptoms were not reported unless their presence or absence was documented in the medical records. Notably, the Childhood Brain Tumor Consortium cohort17 reported a higher frequency of headache, nausea and vomiting, and papilloedema in supratentorial tumours than our study, but reported a similar frequency of these symptoms in infratentorial tumours. This difference is probably due to increased imaging availability to our cohort. Because of the vulnerability of the cerebral aqueduct to compression by tumour, posterior fossa tumours often lead to raised intracranial pressure at an early stage. By contrast, supratentorial tumours could present with other symptoms and signs and grow to a large size before they lead to raised intracranial pressure. The availability of CT and MRI allows the latter children with supratentorial tumours to be assessed before the development of raised intracranial pressure. http://oncology.thelancet.com Vol 8 August 2007
Articles
In the meta-analysis, the frequency of change in behaviour or school performance (7% in all brain tumours, and 9% for central tumours) was lower than that reported by many individual studies. Several large cohorts reported a frequency of school difficulties and behavioural changes of 22–72%.38,44,68,98 Lethargy was analysed separately in this study (pooled proportion: 6% for all intracranial tumours, 21% in children with intracranial tumours aged under 4 years, 13% in posterior fossa tumours), which could account for some of the difference. Adults with brain tumours are often not asked about behavioural change, and similar reporting errors probably occur in children.99 The literature search for this analysis identified several deficits in this specialty. Most identified studies were done in one centre. Only four studies reported symptoms and signs from children enrolled in national trials, of which two reported low numbers of patients.39,56,57,85 Many multi-institutional and multinational trials are done in paediatric neuro-oncology, although these studies, while reporting survival, rarely report symptoms and signs. These symptoms and signs would be easy to obtain, and would improve the level of evidence in this area. Patients and their families find the extended symptom interval associated with paediatric brain tumours very distressing,16 and would probably be willing to provide this information if it could aid earlier diagnosis for future children. The increasing involvement of patients and their families in the development of oncology trials could encourage institutions to obtain these data in future protocols. Little published information exists on brain tumour presentation and diagnosis in adolescents and young adults. This is a population who have less parental supervision than younger children, in whom mood disturbance and behavioural change are common and in which individuals might be less willing to engage with health-care providers. Adolescents and young adults show a different tumour epidemiology to children and often have disturbances of growth and puberty.100 Therefore, we caution the application of information obtained from across all age groups to adolescents and young adults and, just as infants and young children are often reported separately from older children, they should be regarded as a separate age group. No published data exist on symptom progression in children with CNS tumours. If the diagnostic pathway is to be improved, more information is needed on the progression of symptoms and signs in children, from symptom onset to diagnosis and the factors that prompt imaging. Such studies would have to be retrospective, should obtain information directly from patients or carers, and will be relatively expensive; however, only by understanding how symptoms and signs of CNS tumours develop and the factors that lead to a request for imaging can more rapid diagnosis be achieved. This analysis shows both the heterogeneity of childhood CNS tumour presentation and the importance of tumour http://oncology.thelancet.com Vol 8 August 2007
location, age, and neurofibromatosis status in presentation. By ranking symptoms and signs and reporting by age and tumour location, this study focuses on the associative features in a hierarchical way that could be useful for training. Symptoms and signs of raised intracranial pressure occur in less than 50% of all children with intracranial tumours. Motor system abnormalities, especially abnormalities of gait and coordination, are common with all tumour types. Eye signs are common in all intracranial tumour types. Macrocephaly is common in children under 4 years who have intracranial tumours. Weight loss occurs with all tumour types, growth failure with central tumours, and precocious puberty in children with neurofibromatosis and intracranial tumours. Assessment of any child who presents with symptoms and signs that could result from a CNS tumour should include a thorough visual and motor system examination, assessment of growth (including head circumference in children under 4 years), and pubertal status. Specific multiple symptoms and signs (eg, in the combinations shown in figure 4), should alert the clinician to the possibility of a CNS tumour. Contributors DW, JC, CK, RG, and SW devised the study. JC provided statistical advice. SW undertook the data collection, data analysis, and drafted the manuscript. All authors commented on manuscript drafts and contributed to the final version. Conflicts of interest The authors declared no conflicts of interest. Acknowledgments SW was funded by a research grant from The Big Lottery Fund in conjunction with The Samantha Dickson Brain Tumour Trust. References 1 Office for National Statistics. Deaths by age, sex and underlying cause, 2003 registrations. Health Stat Q; quarterly 22. 2 UNICEF. Monitoring the situation of children and women. http://www.childinfo.org/ (accessed April 26, 2007). 3 Young G, Toretsky J, Campbell A, Eskenazi A. Recognition of common childhood malignancies. Am Fam Physician 2000; 61: 2144–54. 4 NICE. Referral guidelines for suspected cancer in adults and children. http://www.nice.org.uk (accessed April 26, 2007). 5 Cancer Research UK. News & resources. http://info. cancerresearchuk.org/cancerstats/ (accessed April 26, 2007). 6 Davison K, Ramsey M. The epidemiology of acute meningitis in children in England and Wales. Arch Dis Child 2003; 88: 662–64. 7 EURODIAB ACE Study Group. Variation and trends in incidence of childhood diabetes in Europe. Lancet 2000; 355: 873–76. 8 Aarsen F, Paquier P, Reddingius R, et al. Functional outcome after low-grade astrocytoma treatment in childhood. Cancer 2006; 106: 396–402. 9 Ilveskoski I, Pihko H, Wiklund T, et al. Neuropsychologic late effects in children with malignant brain tumors treated with surgery, radiotherapy and “8 in 1” chemotherapy. Neuropediatrics 1996; 27: 124–29. 10 Macedoni-Luksic M, Jereb B, Todorovski L. Long-term sequelae in children treated for brain tumors: impairments, disability, and handicap. Pediatr Hematol Oncol 2003; 20: 89–101. 11 Kennedy C, Leyland K. Comparison of screening instruments for disability and emotional/behavioral disorders with a generic measure of health-related quality of life in survivors of childhood brain tumors. Int J Cancer 1999; 12 (suppl): 106–11. 12 Martinez-Climent J, Sanchez V, Menor C, et al. Scale for assessing quality of life of children survivors of cranial posterior fossa tumors. J Neurooncol 1994; 22: 67–76.
693
Articles
13 14
15
16 17
18 19
20 21
22
23
24 25 26 27 28 29
30
31
32
33
34
35
36 37
38 39
40
694
Hansard. Brain cancer (children). London: House of Commons, 2006: column 825. Cancer: clinical features and management. Section 6.6 in: Warrell D, Cox T, Firth J, Benz E, eds. Oxford textbook of medicine, 4th edn. Oxford: Oxford University Press, 2004. Haslam RHA. Brain tumors in children. In: Nelson W, Behrman R, Kliegman R, Arvin A, eds. Textbook of pediatrics, 15th edn. Philadelphia: Harcourt Brace and Company, 1996: 1731–35. Dixon-Woods M, Findlay M, Young B, et al. Parents’ accounts of obtaining a diagnosis of childhood cancer. Lancet 2001; 357: 670–74. Childhood Brain Tumor Consortium. The epidemiology of headache among children with brain tumor. J Neurooncol 1991; 10: 31–46. Lipsey M, Wilson D. Practical meta-analysis. Thousand Oaks, CA, USA: SAGE publications, 2001. Czyzyk E, Jowiak S, Roszkowski M, Schwartz R. Optic pathway gliomas in children with and without neurofibromatosis 1. J Child Neurol 2003; 18: 471–78. Tewari M, Sharma B, Mahajan R, et al. Supratentorial tumours in infants. Child Nerv Syst 1994; 10: 172–75. Balestrini M, Micheli R, Giordano L, et al. Brain tumors with symptomatic onset in the first two years of life. Childs Nerv Syst 1994; 10: 104–10. Haddad S, Menezes A, Bell W, et al. Brain tumors occurring before 1 year of age: a retrospective reviews of 22 cases in an 11-year period (1977–1987). Neurosurgery 1991; 29: 8–13. Trujillo-Maldonado A, Davila-Gutierrez G, Escanero-Salazar A, et al. [Brain tumors in nursing infants]. Bol Med Hosp Infant Mex 1991; 48: 807–13. Lund A, Skovby F. Optic gliomas in children with neurofibromatosis type 1. Eur J Pediatr 1991; 150: 835–38. Haddad S, Moore S, Menezes A, VanGilder J. Ganglioglioma: 13 years of experience. Neurosurgery 1992; 31: 171–78. Smith N, Carli M, Hanieh A, et al. Gangliogliomas in childhood. Childs Nerv Syst 1992; 8: 258–62. Vandertop W, Hoffman H, Drake J, et al. Focal midbrain tumors in children. Neurosurgery 1992; 31: 186–94. Packer R, Parilongo G, Johnson D, et al. Choroid plexus carcinoma of childhood. Cancer 1991; 69: 580–85. Hope J, Armstrong D, Babyn P, et al. Primary meningeal tumors in children: correlation of clinical and CT findings with histologic type and prognosis. Am J Neuroradiol 1992; 13: 1353–64. Reed U, Rosemberg S, Gherpelli J, et al. Brain tumors in the first two years of life: a review of forty cases. Pediatr Neurosurg 1993; 19: 180–85. Hoppe-Hirsch E, Hirsch J. [Malignant tumors of the cerebral hemisphere in children. A series of 64 cases]. Arch Fr Pediatr 1993; 50: 403–07. Hung P, Wang H, Chou M, et al. Intracranial meningiomas in childhood. Zhonghua Min Guo Xiao Er Ke Yi Xue Hui 1994; 35: 495–501. Kuenzle C, Weissert M, Roulet E, et al. Follow-up of optic pathway gliomas in children with neurofibromatosis. Neuropediatrics 1994; 25: 295–300. Gordon G, Wallace S, Neal J. Intracranial tumours during the first two years of life: presenting features. Arch Dis Child 1995; 73: 345–47. Robertson P, Muraszko K, Brunberg J, et al. Pediatric midbrain tumors: a benign subgroup of brainstem gliomas. Pediatr Neurosurg 1995; 22: 65–73. Zhou B, Li J, Luo Z. [Clinicopathologic analysis of ependymoma in childhood]. Zhonghua Bing Li Xue Za Zhi 1995; 24: 369–71. Molloy PT, Bilaniuk LT, Vaughan SN, et al. Brainstem tumors in patients with neurofibromatosis type 1: a distinct clinical entity. Neurology 1995; 45: 1897–902. Edgeworth J, Bullock P, Bailey A, et al. Why are brain tumours still being missed? Arch Dis Child 1996; 74: 148–51. Kaplan A, Albright A, Zimmerman R, et al. Brainstem gliomas in children. A Children’s Cancer Group review of 119 cases. Pediatr Neurosurg 1996; 24: 185–92. Chintagumpala M, Armstrong D, Miki S, et al. Mixed neuronal-glial tumors (gangliogliomas) in children. Pediatr Neurosurg 1996; 24: 306–13.
41 42
43 44 45
46
47
48 49 50 51
52 53
54
55
56
57
58 59
60 61
62 63
64 65 66
67
68
Dirks P, Harris L, Hoffman H, et al. Supratentorial primative neuroectodermal tumours in children. J Neurooncol 1996; 29: 75–84. Constantino S, Houten J, Miller D, et al. Intramedullary spinal cord tumors in children under the age of 3 years. J Neurosurg 1996; 85: 1036–43. Parker A, Robinson R, Bullock P. Difficulties in diagnosing intrinsic spinal cord tumours. Arch Dis Child 1996; 75: 204–07. Bognar L. Brain tumours during the first year of life. Ann N Y Acad Sci 1997; 824: 148–55. Chitnavis B, Phipps K, Harkness W, Hayward R. Intrinsic brainstem tumours in childhood: a report of 35 children followed for a minimum of 5 years. Br J Neurosurg 1997; 11: 206–09. Behnke J, Christen H-J, Bruck W, Marakis E. Intra-axial endophytic tumors in the pons and/or medulla oblongata. I. Symptoms, neuroradiological findings, and histopathology in 30 children. Childs Nerv Syst 1997; 13: 122–34. Johnson J, Hariharan S, Berman J, et al. Clinical outcome of pediatric gangliogliomas: ninety-nine cases over 20 years. Pediatr Neurosurg 1997; 27: 203–07. Erdincler P, Lena G, Sarioglu A, et al. Intracranial meningiomas in children: review of 29 cases. Surg Neurol 1997; 49: 136–41. Lin I, Shu S, Chu H, Chi C. Primary intracranial germ-cell tumor in children. Chinese Med J 1997; 60: 259–64. Bilaniuk L, Molloy P, Zimmerman R, et al. Neurofibromatosis type 1: brain stem tumours. Neuroradiology 1997; 39: 642–53. Przybylski G, Albright A, Martinez A. Spinal cord astrocytomas: long-term results comparing treatments in children. Childs Nerv Syst 1997; 13: 375–82. Furuta T, Tabuchi A, Adachi Y, et al. Primary brain tumors in children under age 3 years. Brain Tumor Pathol 1998; 15: 7–12. Jovani Casano C, Canete Nieto A, Bermudez Cortes M, et al. [Central nervous system tumors in children less than three years of age]. An Esp Pediatr 1998; 49: 151–56. Chung S-K, Wang K-C, Nam D-H, Cho B-K. Brain tumor in the first year of life: a single institute study. J Korean Med Sci 1998; 13: 65–70. Young Poussaint T, Kowal J, Barnes P, et al. Tectal tumours of childhood: clinical and imaging follow-up. Am J Neuroradiol 1998; 19: 977–83. Berger C, Thiesse P, Lellouch-Tubian A, et al. Choroid plexus carcinomas in childhood: clinical features and prognostic factors. Neurosurgery 1998; 42: 470–75. Comi A, Backstrom J, Burger P, Duffner P. Clinical and neuroradiologic findings in infants with intracranial ependymomas. Pediatric Oncology Group. Pediatr Neurol 1998; 18: 23–29. Bouffet E, Pierre-Kahn A, Marchal J, et al. Prognostic factors in pediatric spinal cord astrocytoma. Cancer 1998; 83: 2391–99. Lonjon M, Goh K, Epstein F. Intramedullary spinal cord ependymomas in children: treatment, results and follow-up. Pediatr Neurosurg 1998; 29: 178–83. Keene D, Hsu E, Ventureyra E. Brain tumors in childhood and adolescence. Pediatr Neurol 1999; 20: 198–203. Sala F, Cloarusso E, Mazza C, et al. Brain tumors in children under 3 years of age. Recent experience (1987–1997) in 39 patients. Pediatr Neurosurg 1999; 31: 16–26. Kansal S, Jindal A, Mahapatra A. Brain stem glioma—a study of 111 patients. Indian J Cancer 1999; 36: 99–108. Grant G, Avellino A, Loeser J, et al. Management of intrinsic gliomas of the tectal plate in children. A ten-year review. Pediatr Neurosurg 1999; 31: 170–76. Zuccaro G, Sosa F, Cuccia V, et al. Lateral ventricle tumors in children: a series of 54 cases. Childs Nerv Syst 1999; 15: 774–85. Drummond K, Rosenfeld J. Pineal region tumours in childhood. A 30-year experience. Childs Nerv Syst 1999; 15: 119–26. Yang H, Nam D, Wang K, et al. Supratentorial primitive neuroectodermal tumor in children: clinical features, treatment outcome and prognostic factors. Childs Nerv Syst 1999; 15: 377–83. Young Poussaint T, Yousef N, Barnes P, et al. Cervicomedullary astrocytomas of childhood: clinical and imaging follow-up. Pediatr Radiol 1999; 29: 662–68. Thulesius H, Pola J, Hakansson A. Diagnostic delay in pediatric malignancies—a population-based study. Acta Oncol 2000; 39: 873–76.
http://oncology.thelancet.com Vol 8 August 2007
Articles
69
70
71
72 73
74
75
76
77
78
79
80
81 82 83
84 85
McEvoy A, Harding B, Phipps K, et al. Management of choroid plexus tumours in children: 20 years experience at a single neurosurgical centre. Pediatr Neurosurg 2000; 32: 192–99. Akyuz C, Emir S, Akalan N, et al. Intracranial ependymomas in childhood—a retrospective review of sixty-two children. Acta Oncol 2000; 39: 97–100. Amirjamshidi A, Mehrazin M, Abbassioun A. Meningiomas of the central nervous system occurring below the age of 17: report of 24 cases not associated with neurofibromatosis and review of literature. Childs Nerv Syst 2000; 16: 406–16. Hamelin N, Becquet P, Pierre-Kahn A, Dufier J-L. Clinical features of gliomas of the chiasma. J Fr Ophthalmol 2000; 23: 694–98. Mucic-Pucic B, Cvitanovic-Sojat L, Hajnzic T, Mataija M. Neurologic symptoms as first signs of brain tumour in children. Acta Clinica Croatica 2001; 40: 27–30. Due-Tonnessen B, Helseth E, Skullerud K, Lundar T. Choroid plexus tumors in children and young adults: report of 16 consecutive cases. Childs Nerv Syst 2001; 17: 252–56. Halperin E, Watson D, George S. Duration of symptoms prior to diagnosis is related inversely to presenting disease stage in children with medulloblastoma. Cancer 2001; 91: 1444–50. Nishio S, Morioka T, Taweshita I, Fukui M. Nerve cell tumours of the cerebrum: variable clinical and pathological manifestations. J Clin Neurosci 2001; 8: 225–30. Rohrer T, Gassman K, Buchfelder M, et al. [Clinical symptoms in 35 children and adolescents with craniopharyngeoma at the time of diagnosis]. Klin Padiatr 2002; 214: 285–90. Dobrovoljac M, Hengartner H, Boltshauser E, Grotzer MA. Delay in the diagnosis of paediatric brain tumours. Eur J Pediatr 2002; 161: 663–67. Mehta V, Chapman A, McNeely P, et al. Latency between symptom onset and diagnosis of pediatric brain tumors: an eastern Canadian geographic study. Neurosurgery 2002; 51: 365–72. Im S, Chung C, Cho B, et al. Intracranial ganglioglioma: preoperative characteristics and oncologic outcome after surgery. J Neurooncol 2002; 59: 173–83. Martinez-Lage J, Perez-Espejo M, Esteban J, Paza M. Thalamic tumors: clinical presentation. Childs Nerv Syst 2002; 18: 405–11. Zwerdling T, Dothage J. Meningiomas in children and adolescents. J Pediatr Hematol Oncol 2002; 24: 199–204. Rivera-Luna R, Medina-Sanson A, Leal-Leal C, et al. Brain tumors in children under 1 year of age: emphasis on the relationship of prognostic factors. Childs Nerv Syst 2003; 19: 311–14. Young H, Johnston H. Intracranial tumors in infants. J Child Neurol 2004; 19: 424–30. Dai A, Backstrom J, Burger P, Duffner P. Supratentorial primitive neuroectodermal tumors of infancy: clinical and radiologic findings. Pediatr Neurol 2003; 29: 430–34.
http://oncology.thelancet.com Vol 8 August 2007
Alston R, Newton R, Kelsey A, et al. Childhood medulloblastoma in northwest England 1954 to 1997: incidence and survival. Dev Med Child Neurol 2003; 45: 308–14. 87 Guillamo J, Creange A, Kalifa C, et al. Prognostic factors of central nervous system tumours in neurofibromatosis 1 (NF1): a retrospective study of 104 patients. Brain 2003; 126: 152–60. 88 King A, Listernick R, Charrow J, et al. Optic pathway gliomas in neurofibromatosis type 1: the effect of presenting symptoms on outcome. Am J Med Genet 2003; 122: 95–99. 89 Wanyoike P. Posterior cranial fossa tumours in children at Kenyatta National Hospital, Nairobi. East Afr Med J 2004; 81: 258–60. 90 Sharma M, Ralte A, Gaekwad S, et al. Subependymal giant cell astrocytoma: a clinicopathological study of 23 cases with special emphasis on proliferative markers and expression of p53 and retinoblastoma gene proteins. Pathology 2004; 36: 139–44. 91 Thiagalingam S, Flaherty M, Billson F, North K. Neurofibromatosis type 1 and optic pathway gliomas: follow-up of 54 patients. Ophthalmology 2004; 111: 568–77. 92 Perek D, Drogosiewicz M, Dembowska-Baginska B, et al. Diagnostic problems in children with primary brain tumours treated in Children’s Memorial Health Institute. Pediatria Polska 2005; 80: 29–36. 93 Pollack IF. Brain tumors in children. N Engl J Med 1994; 331: 1500–07. 94 Moher D, Cook D, Eastwood S, et al. Improving the quality of reports of meta-analyses of randomised controlled trials: the QUOROM statement. Quality of reporting of meta-analyses. Lancet 1999; 354: 1896–900. 95 Phillips B, Ball C, Sackett D, et al. Oxford Centre for Evidence-based Medicine levels of evidence. 2001. http://www.cebm.net/index. aspx?o=1025 (accessed April 26, 2007). 96 Linet M, Ries L, Smith M, et al. Cancer surveillance series: recent trends in childhood cancer incidence and mortality in the United States. J Natl Cancer Inst 1999; 91: 1051–58. 97 Rickert C, Werner, P. Epidemiology of central nervous system tumors in childhood and adolescence based on the new WHO classification. Childs Nerv Syst 2001; 17: 503–11. 98 Wilne S, Ferris R, Nathwani A, Kennedy C. The presenting features of brain tumours: a review of 200 cases. Arch Dis Child 2006; 91: 502–06. 99 Davies E, Clarke C. Early symptoms of brain tumours. J Neurol Neurosurg Psychiatry 2004; 57: 1205–06. 100 Levy A. Brain tumors in children: evaluation and management. Curr Prob Pediatr Adolesc Health Care 2005; 35: 230–45. 86
695