J Neurooncol (2009) 94:51–56 DOI 10.1007/s11060-009-9799-2
CLINICAL STUDY - PATIENT STUDY
Medulloblastoma in children under the age of three years: a retrospective Canadian review Donna L. Johnston Æ Daniel Keene Æ Ute Bartels Æ Anne-Sophie Carret Æ Bruce Crooks Æ David D. Eisenstat Æ Chris Fryer Æ Lucie Lafay-Cousin Æ Valerie Larouche Æ Albert Moghrabi Æ Beverly Wilson Æ Shayna Zelcer Æ Mariana Silva Æ Josee Brossard Æ Eric Bouffet
Received: 4 September 2008 / Accepted: 19 January 2009 / Published online: 31 January 2009 Ó Springer Science+Business Media, LLC. 2009
Abstract Children under the age of 3 with medulloblastoma have an inferior survival to older children with this disease. This study reviewed the incidence, characteristics, therapy, and outcome of children less than 36 months of age diagnosed with medulloblastoma from 1990 to 2005 in Canada. Ninety-six cases were identified with a median age at diagnosis of 19.5 months. Forty-seven percent of patients had a complete resection, 25% a 90–95% near complete resection, 20% an incomplete (10–90%) resection, and 3% biopsy only. Therapy consisted of chemotherapy (90%), high dose chemotherapy with stem cell rescue (13%), and radiation therapy (21%). The
median survival time was 45 ± 13.82 months. There was no significant difference in survival when comparing patients with \90% resection versus [90% resection, nor when comparing the presence of metastases versus their absence. There was a significant increase in survival time in patients who received radiation therapy compared to those who were not treated with this modality, as well as for those who were over 18 months at diagnosis compared to those under 18 months.
D. L. Johnston (&) Division of Oncology, Children’s Hospital of Eastern Ontario, 401 Smyth Road, K1H 8L1 Ottawa, Ontario, Canada e-mail:
[email protected]
L. Lafay-Cousin Division of Oncology, Alberta Children’s Hospital, Calgary, Canada
D. Keene Division of Neurology, Children’s Hospital of Eastern Ontario, Ottawa, Canada U. Bartels E. Bouffet Division of Oncology, The Hospital for Sick Children, Toronto, Canada A.-S. Carret Division of Oncology, Montreal Children’s Hospital, Montreal, Canada B. Crooks Division of Oncology, IWK, Halifax, Canada
Keywords
Medulloblastoma Infants
V. Larouche Division of Oncology, Centre Hospitalier Universitaire de Quebec, Quebec, Canada A. Moghrabi Division of Oncology, Ste Justine Hospital, Montreal, Canada B. Wilson Division of Oncology, Stollery Children’s Hospital, Edmonton, Canada S. Zelcer Division of Oncology, Children’s Hospital of Western Ontario, London, Canada
D. D. Eisenstat Cancer Care Manitoba, Winnipeg, Canada
M. Silva Division of Oncology, Kingston General Hospital, Kingston, Canada
C. Fryer Division of Oncology, BC Children’s Hospital, Vancouver, Canada
J. Brossard Division of Oncology, Centre Hospitalier Universitaire de Sherbrooke, Sherbrooke, Canada
123
52
Introduction Medulloblastoma is the most common malignant childhood central nervous system (CNS) tumor, accounting for 20% of all CNS tumors in patients under the age of 18 years, with peak incidences at the ages of 3–4 years and 8–9 years [1]. In children under 3 years of age, the most common tumor types are glioma, medulloblastoma-PNET and choroid plexus tumor [2]. Children under the age of 3 years diagnosed with medulloblastoma have an inferior outcome compared to older children, in large part due to reluctance to use radiation therapy in this young age group. Two year overall survival ranges for young children with medulloblastoma range from 46 to 70% and 5 year overall survival ranges from 13 to 60% [3–9]. Previous studies of medulloblastoma in infants were mostly prospective studies with strict selection criteria, such as staging requirements, minimum Lansky performance status, and stringent deadlines to start treatment, thus may have some bias and not truly reflect the natural history of this disease in early childhood. In addition they had differing age ranges for infants and did not have large numbers of patients included in their analysis. In this population based Canadian study, we examined the incidence, characteristics, therapy, and outcome of children less than 36 months of age diagnosed with medulloblastoma from 1990 to 2005.
Methods Canada has a unique national universal health care system resulting in regionalization of complex health care. Pediatric oncology care in Canada is provided in one of 17 centers. The Canadian Pediatric Brain Tumor Consortium (CPBTC) is a group of clinicians representing all centers in Canada that treat children with brain tumors. Through the CPBTC, a standardized questionnaire was distributed collecting data on all children less than 36 months of age diagnosed with a brain tumor between the years 1990 and 2005. 1990 was chosen as the starting date for data collection as it represents the point in time when magnetic resonance imaging (MRI) was available at all centers in the CPBTC. The questionnaire was completed by all participating centers of the CPBTC except Newfoundland, and a data bank was established. Cases of histologically confirmed medulloblastoma were extracted from this data bank and the following characteristics were obtained: duration of symptoms prior to diagnosis, treatment duration, follow up time, year of diagnosis, age at diagnosis, sex, presenting symptoms, location of tumor, degree of resection, presence of disseminated disease at diagnosis, treatment, and occurrence of event. Degree of resection was classified as: none, biopsy (\10%
123
J Neurooncol (2009) 94:51–56
resection), incomplete (10–90% resection), near complete (90–95% resection), or complete ([95% resection). A mean age adjusted incidence rate per 100,000 children years was calculated using the Statistics Canada population data-base for the year 2000. Chi square analysis was used for nonparametric analysis, whereas the T test was used for parametric analysis. Survival was analyzed using Kaplan– Meier analysis with the use of the log rank test to test for significance between the survival curves. Significance was defined as a P-value of less than 0.05.
Results The total number of children less than 36 months diagnosed with a brain tumor in the data bank was 579. The number of cases of medulloblastoma among these tumors was 96 (16.6%) with a mean age adjusted incidence of 0.58 ± 0.3 (95% CI 0.42–0.74) cases per 100,000 children years. Fifty-three percent of cases were male and 45% female (2% missing data). The age grouping at diagnosis showed that most cases occurred in the 25–36 month age range (Table 1) with a median age at diagnosis of 20 months. Grouping patients by year of diagnosis showed a relatively equal distribution with 30% diagnosed between 1990 and 1995, 42% diagnosed between 1996 and 2000, and 28% diagnosed between 2001 and 2005. The presenting symptoms showed that the majority of children presented with vomiting (56%) and gait disorders (53%), followed by enlarging head circumference (19%) and failure to thrive (10%). The median duration of symptoms prior to diagnosis was 5 weeks with a range of 0–52 weeks. The location of the tumor was in the cerebellum in 99% and in the brain stem in 1 case (1%). The degree of resection obtained showed that 47% of patients had a complete resection, while 72% had a greater than 90% resection. Only 3% of patients had a biopsy only. At diagnosis, metastatic disease (either positive cerebral spinal fluid (CSF) or positive imaging) was present in 43% of patients (Table 2). Among the 96 patients, 9 (9%) received no further therapy following surgery, and these patients were equally distributed in terms of age (one neonate, three between ages of 1–12 months, two between ages Table 1 Age grouping at the time of diagnosis Age
Number of cases
Percentage (%)
Neonate (\1 month)
3
3.1
1–6 months
7
7.3
7–12 months
14
14.6
13–24 months
34
35.4
25–36 months
38
39.6
J Neurooncol (2009) 94:51–56
53
Table 2 Presence of disseminated disease at diagnosis Location of disseminated disease
Number
Percentage (%)
Cerebrospinal fluid
Imaging
Not available
Not available
3
3
Not available
Positive
11
11
Not available
Negative
10
10
Positive
Not available
1
1
Positive
Positive
17
18
Positive
Negative
6
6
Negative Negative
Not available Positive
2 6
2 6
Negative
Negative
39
41
41
43
Overall metastatic disease
of 13–24 months and three between ages of 24–36 months), degree of resection (one autopsy diagnosis, one biopsy, three incomplete, two near complete, two complete), year of diagnosis, presence of metastatic disease, and reason for no treatment given (one diagnosis at autopsy, one intra-operative death, one death prior to starting chemotherapy, two parental refusal, and 4 no reason given). Of the remaining 87 patients, 86 received further therapy with chemotherapy; 12 had high dose chemotherapy with autologous stem cell rescue, 20 received radiation therapy (seven to the tumor bed, one to the whole brain, and 12 to the cranio-spinal axis), and one received radiation therapy only. Of these 20 who received radiation therapy, 14 were alive at the time of the survey. Sixty-seven patients received only chemotherapy and 20 (30.3%) were alive at the time of the survey. Nineteen received chemotherapy and radiation therapy, and 14 (73.7%) were alive at the time of the survey; seven received chemotherapy and focal radiation and five (71.4%) were alive. Sixty-six percent of patients developed recurrent disease with a median time to recurrence of 27 ± 39.7 months. Of these patients, 16 subsequently received chemotherapy, seven additional surgery, 16 radiation therapy and, 31 received only supportive care. Information was not available for seven cases. Seven of these patients with recurrent disease were alive at the time of the survey, of which five had received radiation therapy at the time of their recurrence. The mean follow up time was 53.0 ± 49.5 months with a median follow up time of 35.5 months. The median event Table 3 Comparison of outcome and therapies during the time periods of the survey
Alive at time of survey
free period was 31 ± 3.6 (95%CI 24.0–37. 9) months. Overall 57 ± 5% cases remained event free 2 years after diagnosis and 30 ± 5% continued to be event free at 5 years post diagnosis. No significant change in the rate of patients remaining event free occurred over the time period of the survey (1990–1995: 50 ± 10% at 2 years and 28 ± 9% at 5 years; 1995–2000: 64 ± 7% at 2 years and 28 ± 7% at 5 years; and 2001–2005: 65 ± 9% at 2 years and 35 ± 10% at 5 years) (Table 3). At the time of the survey, 35% of patients were alive. The overall median survival time was 45 ± 13.8 (95% CI 17.9–72.1) months (Fig. 1). The 2 year survival was 67.5 ± 5% and the 5 year overall survival was 45.7 ± 5%. Survival rates did not vary significantly over time. From 1990 to 1995, the 2 year survival rate was 55 ± 9% and the 5 year survival was 40.9 ± 9%. The 2 year survival rate from 1996 to 2000 was 70 ± 7% and the 5 year survival rate was 47 ± 8%. From 2001 to 2005, the 2 year survival rate was 67.5 ± 5% with a 5 year survival rate of 47 ± 15%. When comparisons were made based on being alive at the time of the survey, the significant predictors were the use of radiation therapy and the use of chemotherapy with radiation therapy (Table 4). The year of diagnosis, age at diagnosis, sex, use of stem cell transplant, and degree of resection, did not predict survival at the time of the survey. The survival curves comparing the presence or absence of metastatic disease show no significant difference (Fig. 2). When comparing children less than 18 months to those 18–36 months, the overall survival at 2 and 5 years, was significantly better in the over 18 month age group (2 year survival: 81 ± 5% vs 45 ± 8%; 5 year survival: 52 ± 7% vs. 33 ± 8%). As well, patients who were less than 18 months of age received radiation therapy significantly less, were significantly more often female, and had significantly less dissemination of disease (Table 5). When comparing these children based on receipt of radiation therapy and survival, children less than 18 months of age who received radiation therapy (dose 5,000–5,400 cGy) had a mean survival of 67.6 ± 17.3 months in those who did not receive radiation. In contrast, those over 18 months who received radiation therapy (dose 1,800–5,600 cGy) had a mean survival of 128.45 ± 19.72 months compared to 76.21 ± 11.5 months in those who did not receive radiation therapy. 1990–1995 (n = 29)
1995–2000 (n = 40)
2000–2005 (n = 27) 14 (51.9%)
10 (34.5%)
10 (25%)
No treatment
4 (13.8%)
2 (5%)
3 (11%)
Chemotherapy
24 (82.7%)
38 (95%)
24 (88%)
2 (5%)
10 (37%)
6 (15%)
7 (26%)
High dose chemotherapy (SCT) CSI Radiation
0 7 (24%)
123
54
J Neurooncol (2009) 94:51–56
Table 4 Comparison of alive versus not alive for different factors for infants with medulloblastoma
Year of Diagnosis
Age at Diagnosis
Sex Radiation therapy
Not alive
Alive
1990–1995
19
10
1996–2000
30
10
2001–2005
13
14
3
0
Neonate
15
6
13–24 months
22
12
25–36 months Male
22 30
16 21
0.440
Female
30
13
0.188
None
46
20
6
14
47
27
5
7
Less than 90%
18
8
90% or greater
43
26
Alone
14
20
5
46
None Received
Degree of resection Chemotherapy
0.078
1–12 months
Received Stem cell transplant
P-value
With radiation
0.014 0.15 0.351 0.001
Discussion This study, examining 96 children under the age of 36 months with medulloblastoma, is one of the largest in the literature. We found a 2 year overall survival of 67% which is also one of the highest when compared to
Fig. 1 Survival curve for young children with medulloblastoma
123
published reports where the range is 46–70% [3, 4, 6, 7, 9]. The 5 year overall survival in this study of 45%, is similar to the 13–60% reported in the literature [3, 7, 8]. Thus, this large study found that the outcome for Canadian children less than 36 months diagnosed with medulloblastoma between 1990 and 2005 is similar to or better than previously reported outcomes, and this study had no selection bias. This is somewhat less than reassuring since it indicates that we have not made much progress in improving survival in these children in the MRI era despite improved imaging and staging prior to treatment delivery. The presentation of infants with medulloblastoma in this series showed that the majority presented with vomiting and gait disorder, similar to what is observed in older children [1]. In infants, it has been demonstrated that these patients often initially present with intermittent vomiting, macrocephaly, and inability to elevate the eyes [1]. In our group of patients, 19% presented with an enlarging head circumference, and we emphasize that this sign needs to be thoroughly investigated, especially if there is progressive macrocrania. The variety of symptoms at presentation highlights the need to thoroughly investigate any child with abnormal symptoms, especially related to possible intracranial pathology. An important finding was that the outcome was not affected by the presence of metastatic disease, degree of resection, or age when grouped into 6–12 month groupings, however, when grouped as more or less than 18 months this was found to be significant as discussed below. Results of
J Neurooncol (2009) 94:51–56
55
Fig. 2 Survival curves for children with medulloblastoma with and without the presence of metastatic disease (P = 0.652)
Table 5 Comparison of under 18 months to 18 months or greater at time of diagnosis
Variable Chemotherapy SCT Radiation Sex Year of diagnosis Degree resection
Under 18 months No
34 (100%)
52 (98.1%)
No
31 (93.9%)
43 (81.1%)
Yes
2 (6.1%)
10 (18.9%)
No
27 (81.8%)
39 (73.6%)
Yes
6 (18.2%)
14 (26.4%)
Male
13 (35.1%)
38 (66.7%)
Female
24 (64.9%)
19 (33.3%)
1996–2000
15 (38.5%)
25 (43.6%)
2001–2005
13 (33.3%)
14 (24.6%)
2 (5.3%)
1 (5.8%)
Biopsy
14 (36.8%)
9 (15.8%)
7 (18.4%)
17 (29.8%)
Complete
15 (39.5%)
30 (52.6%)
n/a ?ve
11 (28.9%) 3 (7.9%)
13 (22.8%) 21 (36.8%)
-ve
24 (63.2%)
23 (40.4%)
n/a
5 (13.2%)
1 (1.8%)
?ve
7 (18.4‘%)
27 (47.4%)
Near complete
Imaging
P-value
1 (1.9%)
Yes
Incomplete
CSF
0
18 months or over
0.609 0.086 0.006 0.003 0.643
0.07
0.006
-ve
26 (68.4%)
29 (50.9%)
0.003
Either CSF or imaging
?ve
8 (20.5%)
33 (57.9%)
0.000
Treated after surgery
No
5 (12.8%)
4 (7%)
Yes
34 (87.4%)
53 (93%)
0.338
123
56
previous studies have shown that a strong positive prognostic indicator for survival was the ability to achieve a gross total resection of the tumor [5, 9, 10]. Older children with medulloblastoma who have residual disease ([1.5 cm2) or metastatic disease at presentation have a worse outcome [1]. In contrast, we found that the presence of metastatic disease and incomplete resection did not affect overall outcome, which is similar to a previous report by Walter et al. [3]. Previous studies have had conflicting findings on the impact of young age on prognosis [3–5, 9, 11]. Using patient groupings of neonate, 1–12 months, 13– 24 months, and 25–36 months, we found that age did not appear to affect outcome among the groups. However, when we compared children more or less than 18 months of age at time of diagnosis, there was a significantly higher survival in the older grouping of children. This suggests that perhaps there is a different disease in these two age groups and that children over 18 months of age behave more like older children with this disease. One factor that may account for this difference is that radiation therapy was significantly more often used for the over 18 month old patients, and this likely affected survival. However, when comparing children less than 18 months who received radiation therapy, their survival was not different from those under 18 months who did not receive radiation therapy. This suggests that perhaps very young children have a biologically different medulloblastoma. Children less than 36 months of age are often severely affected by radiation therapy, and for this reason clinicians are reluctant to use this therapy in very young children with medulloblastoma. We found in this study, that those children who received therapy with radiation had a significantly higher likelihood of being alive at the time of data collection in this study. The large St. Jude Children’s Research Hospital study of children less than 36 months with medulloblastoma treated all 29 patients with radiation therapy, either at progression or after completion of chemotherapy [3]. They did, however, find that all patients lost cognitive function during and after therapy at a rate of –3.9 intelligent quotient points per year, and there was also a significant decline in sensory functions. This study found a 66% 2 year overall survival, similar to our results, in which only 23% of patients who were treated with any adjuvant modality, received radiation therapy. Thus, clinicians, while justified in their reluctance to use radiation therapy in these young patients, should not completely rule out it’s use, since it does confer a survival advantage. However, survival is possible, albeit less likely, without the use of this modality of therapy, as shown in this and previous studies [5–7, 11].
123
J Neurooncol (2009) 94:51–56
Overall, this population based study is one of the largest examining children less than 36 months with medulloblastoma. We found a 2 year survival of 67% for these patients. We also found that survival was not affected by the presence of metastatic disease or degree of resection. There was a significant increase in survival for patients who received radiation therapy, but there was still a good overall survival for those who did not receive this modality of therapy. There was also an increase in survival in patients diagnosed over the age of 18 months, likely secondary to the increased use of radiation in these patients, but perhaps due to a biologically different tumor in older children. Overall, these results highlight the need for improved therapy for these young patients.
References 1. Packer RJ, Cogen P, Vezina G et al (1999) Medulloblastoma: clinical and biologic aspects. Neurooncology 1:232–250. doi: 10.1215/15228517-1-3-232 2. Di Rocco C, Iannelli A, Papacci F (2001) Infantile brain tumors. In: Keating RF, Goodrich JT, Packer RJ (eds) Tumors of the pediatric central nervous system. Thieme, New York, pp 376–397 3. Walter AW, Mulhern RK, Gajjar A et al (1999) Survival and neurodevelopmental outcome of young children with medulloblastoma at St Jude Children’s Research Hospital. J Clin Oncol 17:3720–3728 4. Duffner PK, Horowitz ME, Krischer JP et al (1999) The treatment of malignant brain tumors in infants and very young children: an update of the Pediatric Oncology Group experience. Neurooncology 1:152–161. doi:10.1215/15228517-1-2-152 5. Geyer JR, Sposto R, Jennings M et al (2005) Multiagent chemotherapy and deferred radiotherapy in infants with malignant brain tumors: a report from the Children’s Cancer Group. J Clin Oncol 23:7621–7631. doi:10.1200/JCO.2005.09.095 6. Mason WP, Grovas A, Halpern S et al (1998) Intensive chemotherapy and bone marrow rescue for young children with newly diagnosed malignant brain tumors. J Clin Oncol 16:210–221 7. Ater JL, van Eys J, Woo SY et al (1997) MOPP chemotherapy without irradiation as primary postsurgical therapy for brain tumors in infants and young children. J Neurooncol 32:2430252. doi:10.1023/A:1005744527443 8. Stiller CA, Bunch KJ (1992) Brain and spinal tumors in children aged under two years: incidence and survival in Britain, 1971–85. Br J Cancer 18:S50–S53 9. Duffner PK, Horowitz ME, Krischer JP et al (1993) Postoperative chemotherapy and delayed radiation in children less than three years of age with malignant brain tumors. N Engl J Med 328:1725–1731. doi:10.1056/NEJM199306173282401 10. Grill J, Sainte-Rose C, Jouvet A et al (2005) Treatment of medulloblastoma with postoperatrive chemotherapy alone: an SFOP prospective trial in young children. Lancet Oncol 6: 573–580. doi:10.1016/S1470-2045(05)70252-7 11. Rutkowski S, Bode U, Deinlein F et al (2005) Treatments of early childhood medulloblastoma by postoperative chemotherapy alone. N Engl J Med 352:978–986. doi:10.1056/NEJMoa042176