Doi:10.1016/s1470-2045(06)70867-

Articles
Risk-adapted craniospinal radiotherapy followed by high-dose
chemotherapy and stem-cell rescue in children with newly
diagnosed medulloblastoma (St Jude Medulloblastoma-96):
long-term results from a prospective, multicentre trial
Amar Gajjar, Murali Chintagumpala, David Ashley, Stewart Kellie, Larry E Kun, Thomas E Merchant, Shaio Woo, Greg Wheeler, Valerie Ahern, Matthew J Krasin, Maryam Fouladi, Alberto Broniscer, Robert Krance, Gregory A Hale, Clinton F Stewart, Robert Dauser, Robert A Sanford, Christine Fuller, Ching Lau, James M Boyett, Dana Wallace, Richard J Gilbertson Summary
Background Current treatment for medulloblastoma, which includes postoperative radiotherapy and 1 year of Lancet Oncol 2006; 7: 813–20
chemotherapy, does not cure many children with high-risk disease. We aimed to investigate the eff ectiveness of risk- Published Online
adapted radiotherapy followed by a shortened period of dose-intense chemotherapy in children with medulloblastoma.
Methods After resection, patients were classifi ed as having average-risk medulloblastoma (≤1·5 cm2 residual tumour See Refl ection and Reaction
and no metastatic disease) or high-risk medulloblastoma (>1·5 cm2 residual disease or metastatic disease localised to page 787
neuraxis) medulloblastoma. All patients received risk-adapted craniospinal radiotherapy (23·4 Gy for average-risk Department of Oncology
disease and 36·0–39·6 Gy for high-risk disease) followed by four cycles of cyclophosphamide-based, dose-intensive (Prof A Gajjar MD, M Fouladi MD,
chemotherapy. Patients were assessed regularly for disease status and treatment side-eff ects. The primary endpoint A Broniscer MD, G A Hale MD),
Department of Radiological
was 5-year event-free survival; we also measured overall survival. This study is registered with ClinicalTrials.gov, Sciences (M J Krasin MD,
number NCT00003211.
Prof L E Kun MD,
Prof T E Merchant DO),
Findings Of 134 children with medulloblastoma who underwent treatment (86 average-risk, 48 high-risk), 119 (89%) Department of Biostatistics
completed the planned protocol. No treatment-related deaths occurred. 5-year overall survival was 85% (95% CI 75–94) D Wallace MS), Department of
in patients in the average-risk group and 70% (54–84) in those in the high-risk group (p=0·04); 5-year event-free survival Neurosurgery (Prof R A Sanford
was 83% (73–93) and 70% (55–85), respectively (p=0·046). For the 116 patients whose histology was reviewed centrally, MD), Department of
histological subtype correlated with 5-year event-free survival (p=0·04): 84% (74–95) for classic histology, 77% (49–100) Developmental Neurobiology
(R J Gilbertson MD), Department
for desmoplastic tumours, and 57% (33–80) for large-cell anaplastic tumours.
of Pharmaceutical Sciences
(C Stewart PharmD), and
Interpretation Risk-adapted radiotherapy followed by a shortened schedule of dose-intensive chemotherapy can be used Department of Pathology
(C Fuller MD), St Jude Children’s
to improve the outcome of patients with high-risk medulloblastoma.
Research Hospital, Memphis,
Introduction
TN, USA; Texas Children’s
We postulated that the cure rate for newly diagnosed Hospital, Baylor College of
Primitive neuroectodermal tumours, which include medulloblastoma would be improved by using a dose- Medicine, Houston, TX, USA
cerebellar medulloblastoma, are the most common intense chemotherapeutic regimen, and that an intensifi ed (M Chintagumpala MD,
malignant brain tumours in childhood. Two decades ago, regimen would allow a substantial reduction in the S Woo MD, R Dauser MD,
Prof R Krance MD, C Lau MD);
standard treatment of these tumours consisted of surgery duration of treatment. In current US protocols, Children’s Hospital at
and craniospinal radiotherapy, resulting in 10-year survival
chemotherapy is given for about 12 months after Westmead, Sydney, NSW,
of 45%.1 The introduction of chemotherapy before or after completion of radiotherapy. We devised a 4-month dose- Australia (S Kellie MBBS,
radiotherapy greatly improved survival, and more than intense regimen of cyclophosphamide, cisplatin, and V Ahern MBBS); and Children’s
Cancer Centre, Murdoch
70% of children aged 3 years or older with average-risk vincristine that started 6 weeks after completion of risk- Children’s Research Institute,
disease (ie, no evidence of metastasis and residual disease
adapted radiotherapy (protocol St Jude Medulloblastoma Royal Children’s Hospital,
≤1·5 cm2) can now expect to be cured.2 The introduction of [SJMB]-96) and tested the eff ectiveness of this regimen Melbourne, VIC, Australia
eff ective chemotherapy has also helped to reduce the dose prospectively in children with newly diagnosed (D Ashley MBBS, of craniospinal radiotherapy needed to treat average-risk medulloblastoma. High-dose cyclophosphamide has Correspondence to: medulloblastoma. By contrast, survival has remained poor cacy against relapsed and newly diagnosed Dr Amar Gajjar, Division of for children aged 3 years or older who have high-risk medulloblastoma, and the pharmacokinetic and Neuro-Oncology, St Jude medulloblastoma (ie, residual disease >1·5 cm2 or pharmacodynamic properties of alkylating drugs, especially Children’s Research Hospital, metastatic disease), and fewer than 55% of children with cyclophosphamide, favour their use in dose-intense 332 North Lauderdale St, high-risk disease survive for longer than 5 years.3–5 regimens.11–13 Having reported the early outcome and [email protected] Myeloablative chemotherapy has been eff ective in the feasibility of delivering this chemotherapy regimen,14 we treatment of relapsed medulloblastoma, although the high now present long-term outcome data from this prospective mortality associated with this treatment has curtailed its clinical trial and analyse the biological features associated widespread use.6–10 http://oncology.thelancet.com Vol 7 October 2006
Articles
if they had no evidence of metastatic disease (ie, M0), as Patients
confi rmed by gadolinium-enhanced MRI of the head and Between October, 1996, and August, 2003, 134 patients spine, and if lumbar cerebrospinal fl uid assessed at least with newly diagnosed medulloblastoma received treatment 10 days after resection contained no tumour cells; or if at: Texas Children’s Hospital, Baylor College of Medicine, they had no bone metastasis, as confi rmed by bone Houston, TX, USA (n=28); Royal Children’s Hospital, scintigraphy; or if gross total resection was achieved or Melbourne, VIC, Australia (n=14); Children’s Hospital at they had residual disease of 1·5 cm2 or less, as recorded Westmead, Sydney, NSW, Australia (n=12), or St Jude in the operative notes and confi rmed by gadolinium-Children’s Research Hospital, Memphis, TN, USA (n=80). enhanced MRI of the head done within 48 h of resection. Patients aged 3–21 years at the time of diagnosis who had Patients whose operative notes recorded brainstem not received chemotherapy or radiotherapy were eligible invasion but whose MRI showed no visible tumour were for this study. Previous corticosteroid treatment was regarded to have average-risk disease.
allowed. For most patients, treatment started within We defi ned patients as high-risk if they had metastatic 28 days of surgery, the extent of which was defi ned as: disease (as shown by gadolinium-enhanced MRI) in the gross total resection if followed by no evidence of residual head (metastasis [M] stage 2), head and spine (M3), or disease; near-total resection if postoperative MRI showed lumbar spinal fl uid (M1) at least 10 days after resection; or residual disease of 1·5 cm2 or less; and subtotal resection if if they had residual disease of more than 1·5 cm2 (measured 25% or more of the tumour remained. We defi ned the by postoperative gadolinium-enhanced MRI). Patients with extent of resection using the neurosurgeons’ operative bone metastasis or extraneural metastasis were regarded notes and by MRI obtained within 48 h of the surgery. as having M4 disease according to the Chang staging Eligibility criteria also included: normal renal function (ser- system, which would classify these patients as having high- um creatinine ≤110 µmol/L or technetium-99m diethylene- risk disease. However, since the study protocol included triaminepentaacetic acid clearance ≥70 mL/min per m2); stem-cell or bone-marrow rescue, the enrolment of these normal liver function (serum glutamic pyruvic trans patients was judged inappropriate and these patients were aminase ≤1·5 times normal bilirubin and total bilirubin not eligible for the protocol.
≤25·5 µmol/L); normal bone-marrow function (haemo -
globin ≥100 g/L, white-blood-cell count ≥3×109/L, absolute Procedures
neutrophil count ≥1·5×109/L, and platelet count Pathology samples were reviewed by CF at St Jude
≥100×109/L); and a score of 0–3 on the Eastern Cooperative
Children’s Research Hospital. Immunohistochemical Oncology Group (ECOG) performance status scale. Patients with posterior fossa syndrome were eligible for samples were done to determine the activation status of the study even if they did not meet the ECOG criteria at the the wingless (WNT) and sonic hedgehog (SHH) signalling time of starting radiotherapy. However, these patients were pathways, by use of routine methods with primary not eligible for the study if they did not meet the other antibodies against β catenin (CTNNB1, Cell Signaling eligibility criteria (eg, blood variables). The protocol Technology, Danvers, MA, USA) and secreted frizzled-received yearly approval from all institutional ethics boards, related protein (SFRP) 1 (R&D Systems, Minneapolis, MN, and written informed consent was obtained from all USA), respectively.16 Investigators who did the patients, parents, or legal guardians (assent was obtained immunohistochemical analysis were unaware of the risk from older children, when appropriate).
category and outcome data of the patient. Slides were Disease was classifi ed as high risk or average risk, assigned a study number to mask cases before staining. depending on the remaining tumour volume after Staining scores were then submitted to the study surgery and according to a modifi ed Chang staging statisticians (DW, JMB) for independent analysis.
system.15 Patients were placed in the average-risk category Maximum resection of the tumour was attempted in all patients by the senior paediatric neurosurgeon at the participating centre. For patients who had residual disease Panel: High-dose chemotherapy protocol (one cycle)
after resection at their referring institution, a second neurosurgical resection was attempted to achieve gross Day −4—Intravenous doses of 75 mg/m2 cisplatin and 1·5 mg/m2 (maximum 2 mg) vincristine total resection at the participating centre. 31 patients with Days −3 and −2—Intravenous dose of 2 g/m2 cyclophosphamide; continuous infusion of high-risk disease underwent 6 weeks of topotecan treatment as part of a phase II study, after which tumour response was monitored by repeated MRI.17 Patients who Day 0—Infusion of peripheral blood or bone-marrow progenitor cells (peripheral-blood refused topotecan treatment or who were enrolled after the accrual target for topotecan was met received radiotherapy Day +1—Subcutaneous or intravenous dose of 5 μg/kg fi lgrastim per day, until absolute immediately after surgery. Patients with high-risk neutrophil count reaches ≥2 × 109/L on 2 consecutive days after expected nadir of medulloblastoma underwent craniospinal radiotherapy (M0–1, 36 Gy; M2–3, 39·6 Gy) with a three-dimensional Day +6—Intravenous dose of 1·5 mg/m2 vincristine conformal boost to the tumour bed (total dose 55·8 Gy) http://oncology.thelancet.com Vol 7 October 2006
Articles
and, where appropriate, to local sites of metastasis (total Average risk (n=86)
High risk (n=48)
dose 50·4 Gy). Those with average-risk disease received 23·4 Gy craniospinal radiotherapy, 36 Gy radiotherapy to the posterior fossa, and 55·8 Gy to the primary tumour bed. Median duration of radiotherapy was 1·4 months (range 1·3–2·3). After a 6-week rest, all patients began four Age (years)
cycles of high-dose chemotherapy, each of which was followed by stem-cell or bone-marrow rescue (panel). For Ethnic origin
histological review, tumours were defi ned as classic, nodular desmoplastic, or large-cell anaplastic medullo The planned total duration of chemotherapy was 16 weeks, with 4 weeks for every cycle. A new cycle started Extent of resection
once the haemoglobin concentration had reached at least 80 g/L, the platelet count reached at least 75×109/L, and absolute neutrophil count reached at least 1·5×109/L. Additional supportive-care measures were done as described previously.14 In 2000, the protocol was amended Extent of metastasis
to test the hypothesis that amifostine ameliorates cisplatin- induced ototoxic eff ects, the eff ects of which are still In most cases, peripheral-blood stem cells were harvested and cryopreserved after mobilisation with fi lgrastim before Histological category
radiotherapy (average-risk) or after topotecan treatment (high-risk). Patients in whom mobilisation could not be accomplished before beginning craniospinal radiotherapy underwent bone-marrow harvest before starting Unknown (no tumour material available) Immunohistochemical category
time of bone-marrow harvest for rescue to allow all four chemotherapy courses, then patients would undergo a peripheral stem-cell harvest after the fi rst round of chemotherapy. The median number of stem cells harvested was 15·2×106 CD34 cells/kg (range 6–1600×106) with a median of two harvest procedures (range, one–three).
Data are number of patients (%) unless stated otherwise.
During treatment, patients’ disease status and toxic Table 1: Clinical characteristics of enrolled patients
ects were monitored with appropriate laboratory assessments and imaging studies, and toxic eff ects were graded according to the National Cancer Institute’s previously been tumour-free. MRI was used to measure common toxicity criteria.19 Disease status was assessed by disease status in all patients. We defi ned secondary disease MRI of the head and spine. Patients were also monitored as a second malignant process that was not attributed to by routine cytology examination of the cerebrospinal fl uid. the side-eff ects of the protocol. After completion of Complete response was defi ned as disappearance of all treatment, follow-up examinations were done every radiologically discernable tumour and negative cerebro- 3 months for the fi rst 18 months, every 6 months for the spinal fl uid cytology. Partial response was defi ned as at next 5 years, and every year thereafter. We monitored least a 50% decrease in tumour size measured by the sum patients’ physical and disease status and neuroendocrine of the products of the maximum perpendicular diameters and neuropsychological function.
of all measurable lesions and negative cerebrospinal fl uid
cytology (if previously positive). Stable disease was defi ned Statistical analysis
as no reduction in the products of the maximum Our primary endpoint was to estimate 5-year event-free
perpendicular diameters of all measurable lesions, no survival patients with small standard errors. We estimated
progression of any lesion, no new lesions, and no change that 36 average-risk patients and 48 high-risk patients
in cerebrospinal fl uid status.
would be needed to report 5-year estimates of 80% and Progressive disease was classifi ed as disease regrowth at 70%, respectively, with standard errors of less than 7%. By the tumour site of at least 25% of the original tumour, assuming a 5-year event-free survival of 60% in the average-calculated as the product of the perpendicular diameters of risk group on the basis of previous results at St Jude the tumour (as measured by MRI). Recurrence was Children’s Research Hospital,20 we estimated that 36 classifi ed as disease that developed at a new site that had patients in the average-risk group would have a power of http://oncology.thelancet.com Vol 7 October 2006
Articles
80% to detect an improvement of 20% after treatment. The or older. Patients were followed up for a median of 5·0 subsequent amendment to study the eff ect of amifostine years (range 0·43–9·6); 100 (92%) of 109 surviving patients on hearing loss resulted in 86 patients in the average-risk were seen in the past 18 months.
group available for this analysis. The design also included 5-year event-free survival diff ered signifi cantly between rules for early stopping if mortality due to high-dose patients with average-risk disease and those with high-risk chemotherapy exceeded 5%.
disease (p=0·046, fi gure 1), with 83% (73–93) in the average- Event-free survival was measured from the date of risk group and 70% (55–85) in the high-risk group. 3-year enrolment to the date of fi rst occurrence of recurrent or event-free survival was 87% (80–94) in the average-risk progressive disease, death, or second malignant disease group and 75% (63–87) in the high-risk group. Overall after the start of radiotherapy, or to the last follow-up date survival also diff ered signifi cantly between groups; 85% for patients who did not have any of these events. Overall (95% CI 75–94) for the average-risk group and 70% (54–84) survival was measured from the date of enrolment to the for the high-risk group (p=0·04). 5-year event-free survival date of death from any cause or last follow-up. Kaplan- was 66% (48–83) for the 42 patients with metastatic disease. Meier estimates of survival distributions with standard Six patients in the high-risk group who had residual disease errors were calculated with methods used elsewhere.21 but no metastatic disease survived and have no evidence of We used the log-rank test to compare outcome for risk disease at a median follow-up of 8·2 years (range 3·5–8·8).
groups and to investigate associations between the 31 patients in the high-risk group received topotecan, distribution of event-free survival and pathological or and the remaining 17 received radiotherapy after resection. immunohistochemical characteristics of tumours, as Of those given topotecan, only four had progressive disease stratifi ed by risk group. A Cox proportional hazards before undergoing radiotherapy. After further radiotherapy model was used to estimate the hazard ratio for and chemotherapy, three of these four patients remained histological features of tumours in average-risk patients. disease-free for more than 5 years. 5-year event-free survival This study is registered with ClinicalTrials.gov, number was 75% (38–100) for the four patients with progressive NCT00003211. All analyses were done with SAS (version disease during topotecan treatment, 70% (53–88) for the 27 patients who had a complete response, a partial response, or stable disease during topotecan treatment, Role of the funding source
and 71% (33–100) for the 17 patients who did not receive The sponsor of the study had no role in study design, data topotecan. Event-free survival did not diff er between the collection, data analysis, data interpretation, or writing of 31 patients who received topotecan and the 17 patients who the report. The corresponding author had full access to all the data in the study and had fi nal responsibility for the For 44 (92%) patients in the high-risk group, radiotherapy began within 8 weeks of starting topotecan treatment or 28 days after surgery. For patients in the average-risk group, radiotherapy began a median of 27 days (range 21–33) after Median age of the 134 patients enrolled was 7·6 years resection. The median duration of radiotherapy for all (range 3·1–20·2; table 1). Nine (7%) patients were aged 15 years patients was 43 days (range 34–77). Radiotherapy was well tolerated, and we did not record any unexpected grade III or IV toxic eff ects from this treatment.
After radiotherapy, all patients received identical chemo- therapy regimens. 496 courses of chemotherapy were delivered to 129 patients, of whom 119 completed all four cycles. 77 patients in the average-risk group received all four courses (ie, 308 courses). The schedule was modifi ed for six patients in the average-risk group because of insuffi stem cells (n=4), aseptic meningitis (n=1), or myocarditis of unknown cause (n=1). These six patients received 14 courses of chemotherapy. Three patients in the average-risk group were not given high-dose chemotherapy by parental request (n=1) or because of progressive disease (n=2). 42 patients in the high-risk group received all four courses (ie, 168 courses). The schedule was modifi ed for four patients in the high-risk group because of raised creatinine concentrations (n=1), congestive heart failure (n=1), new onset seizures (n=1), or Number at risk
by parental request (n=1). These four patients received six courses of chemotherapy. Two patients in the high-risk group received no chemotherapy because of progressive Figure 1: Event-free survival for average-risk and high-risk patients
disease (n=1) or by parental request (n=1). http://oncology.thelancet.com Vol 7 October 2006
Articles
21 patients in the high-risk group had a complete res- ponse after radiotherapy; six had progressive or recurrent disease at a median of 2·2 years from study enrolment (range 0·96–4·1). 3-year and 5-year event-free survival estimates for this group were 81% (65–92) and 70% (46–93), respectively. The remaining 27 patients in the high-risk group had assessable disease after completion of radio- therapy. Six patients in the average-risk group had assessable disease (ie, had presence of tumour detected by MRI) after radiotherapy and before high-dose chemo- therapy. 18 of 42 patients in the high-risk group and two of 77 in the average-risk group who were assessed 1 month after completion of all four chemotherapy courses had The chemotherapy regimen was tolerated well, with no protocol-related deaths. We recorded 478 events of bone- marrow suppression after 496 courses of high-dose Number at risk
chemotherapy, and 51 events of infection during chemo- therapy in 36 patients. Median time to recover was 13 days (range 8–25) for the absolute neutrophil count to reach more than 0·5×109/L and 17 days (0–36) for the platelet count to reach 20×109/L. Unexpected grade III–IV toxic eff ects were cardiac dysarrythmias or congestive heart failure (n=3), hyperbilirubinaemia (n=2), raised creatinine concentration (n=1), stomatitis (n=9), and increased serum glutamic oxaloacetic transaminase or serum glutamic pyruvic transaminase (n=8). So far, no secondary malignant disease has been recorded in patients enrolled in this trial. One patient developed a second malignant disease (acute lymphoblastic leukaemia) 3 months after completing treat- ment. This patient was included in the outcome analysis. Tumour material from 116 (87%) participants was available for central histological review. The median age of the 22 patients with desmoplastic tumours was 8·2 years (range 3·1–20·1). Of the classic, large-cell anaplastic, and desmoplastic histology groups, fi ve (7%), one (4%), and three (14%) were age 15 years or older. 15 (68%) patients with desmoplastic tumours were white; two (9%) were Number at risk
black. Histological subtype was signifi cantly related to event-free survival after adjustment for clinical stage (p=0·04, fi gure 2A, table 2). Patients in the average-risk group who had large-cell anaplastic tumours had a hazard Figure 2: Event-free survival based on histological subtypes (A) and immunohistochemical subtypes (B)
ratio for treatment failure of 3·9 (95% CI 1·2–12·3) times that of patients with other tumour types. Disease recurrence or progression occurred at the Tumour material was available for immunohistochemical posterior fossa alone in four patients (three average-risk, analysis from 69 (51%) patients. We classifi ed tumours as one high-risk), posterior fossa and brain or spine in three having WNT pathway activation (n=10), SHH pathway patients (one, two), and brain or spine alone in 19 patients activation (n=18), or neither (n=41). No tumour displayed (nine, ten). Bone metastasis developed in one patient in evidence of activation of both pathways. No patients with the high-risk group. For patients with recurrent or WNT pathway mutations had recurrent or progressive dis- progressive disease, median time to relapse was 24 months ease. Since the 5-year event-free survival in patients with (range 4–39) for those at average risk, and 19 months (3–62) SHH pathway activation (65% [43–87]) was similar to that for those at high risk. In 24 patients, relapse occurred after in patients with neither pathway activated (70% [55–85]), completion of all planned chemotherapy. Progressive we pooled these two groups in further analysis (table 2). disease occurred after a median of 18 months (range The stratifi ed exact log-rank test showed that patients with 0·9–54) from completion of chemotherapy. In three WNT pathway abnormalities had better event-free survival patients, disease progressed after completion of than did all other patients (p=0·03, fi gure 2B, table 2).
radiotherapy and before the start of chemotherapy. http://oncology.thelancet.com Vol 7 October 2006
Articles
Histological subtype
Lomustine, cisplatin, vincristine or cyclophosphamide, cisplatin, vincristine Immunohistochemical subtype
Table 2: Event-free survival data by histological and
immunohistochemical subtype
Furthermore, our chemotherapy was completed in Figure 3: Comparison of SJMB-96 protocol for average-risk and high-risk patients with protocols used in
16 weeks, whereas standard treatment for average-risk contemporary national clinical trials
medulloblastoma lasts about 48 weeks (fi gure 3).23,24 In the French study SFOP (table 3),25 patients with Discussion
average-risk medulloblastoma had a lower event-free We have shown that a short, dose-intensive, alkylator-based survival (65%) than did those in US studies (≥79%).22,23 This regimen of chemotherapy after maximum resection and outcome followed reduced-dose (25 Gy) neuraxis radio-risk-adapted radiotherapy in patients with medulloblastoma therapy after chemotherapy for 3 months. By contrast, results in a 5-year event-free survival of 83% (73–93) for patients with average-risk disease in the SIOP study26 those at average risk and 70% (55–85) for those at high received standard-dose neuraxis radiotherapy (36 Gy) after risk. chemotherapy of a similar duration (median 86 days); The ideal clinical trial design to assess the value of our 5-year event-free survival for these patients was 74%. In chemotherapy regimen would be a randomised study that our study, patients in the average-risk group received compares the regimen with surgery and radiotherapy alone. radiotherapy within 4 weeks of resection. The high event- However, various issues preclude such a study. First, free survival recorded with this approach suggests that low neoadjuvant and adjuvant chemotherapy has been the survival results from delayed reduced-dose radiotherapy.25 accepted standard of care of medulloblastoma by all co- If used, reduced-dose neuraxis radiotherapy should begin operative groups for several years, precluding the use of soon after resection.
postoperative radiotherapy alone.3–5,22 Although our data High-risk medulloblastoma, especially if metastatic at could be compared with those for historical controls given the time of diagnosis, confers a very poor prognosis surgery and radiotherapy alone, the value of this comparison (table 3), despite the use of high-dose radiotherapy and is highly questionable. Results with surgery and adjuvant or neoadjuvant chemotherapy. Our 5-year event-radiotherapy alone can be obtained only from studies that free survival rate of 70% for high-risk patients is were done almost two decades ago.1 Advances in imaging, promising. We also report a good survival rate of 66% for surgical technique, and supportive care, would all have had patients with metastatic disease (85% of the high-risk a positive eff ect on outcome.2 Finally, the number of patients cohort). A small study26 (n=15) of patients with metastatic needed for a randomised study would need the eff orts of a medulloblastoma showed a promising 5-year event-free national or international co-operative group. The SJMB survival of 67% (SE 15) using standard-dose chemotherapy, consortium was established to develop and undertake although these results were not reproduced in a randomised initial testing of the feasibility and effi cacy of innovative study by German investigators,5 in which 3-year event-free treatment regimens that could be adopted into larger survival for patients with metastatic disease was 30% (15).
national studies. No other competing treatment protocols
We attribute our results to the early use of high-dose were open at the institutions during the study that might radiotherapy and a short, dose-intense, alkylator-based have resulted in a selection bias in the study population. chemotherapeutic regimen. The craniospinal radiotherapy Attempts to reduce long-term treatment-related morbidity dose is also crucial in achieving improved survival in in average-risk patients have focused on reducing the patients with metastatic disease. Other contemporary neuraxis radiation dose from 36 Gy to 23·4 Gy, and using studies that have used radiotherapy doses to the neuraxis adjuvant chemotherapy to maintain cure rates at more than similar to those used in our study (table 3) have not 75% after 5 years. Our protocol achieved event-free survival achieved durable disease control in patients with high-risk of more than 80% despite reducing the radiation dose to medulloblastoma. Hence, adjuvant dose-intense chemo-the neuraxis, and reducing the total vincristine dose by 75% therapy is an important component of treatment that (from 32 to eight doses) and total cisplatin dose by 50% contributes to the improved survival for high-risk patients. (eight to four doses), relative to doses in standard regimens.22 In addition to our data, high-dose alkylator-based regimens http://oncology.thelancet.com Vol 7 October 2006
Articles
Radiotherapy
Chemotherapy
Planned duration of
5-year event-free Ref
treatment (weeks)
survival (SD)
Average risk
Weekly vincristine during radiotherapy and lomustine, cisplatin, and vincristine (eight cycles at 6-week intervals after radiotherapy) Weekly vincristine during radiotherapy and lomustine, cisplatin, 56 and vincristine; or cyclophosphamide, cisplatin, and vincristine (both eight cycles at 6-week intervals after radiotherapy) Eight drugs in 1 day* or carboplatin and etoposide (both for two 21 High risk
Weekly vincristine during radiotherapy and lomustine, cisplatin, 56 to lumpy tumour and vincristine (eight cycles at 6 week intervals after to 45 Gy) Eight drugs in 1 day* (two cycles) before radiotherapy and eight drugs in 1 day* (eight cycles) after radiotherapy; or weekly vincristine during radiotherapy and vincristine, lomustine, and prednisone (eight cycles) after radiotherapy Vincristine, etoposide, carboplatin or vincristine, etoposide, cyclophosphamide (both two cycles before radiotherapy ) Ifosfamide, cisplatin, etoposide; or cytarabine, methotrexate (two cycles), additional chemotherapy if partial response; or weekly vincristine during radiotherapy and lomustine, cisplatin, and vincristine (eight cycles at 6-week intervals) after radiotherapy *Drugs used are vincristine, methylprednisolone, lomustine, hydroxycarbamide, procarbazine, cisplatin, cyclophosphamide, cytarabine. †3-year event-free survival.
Table 3: Study results of average-risk and high-risk medulloblastoma treated with reduced-dose craniospinal radiotherapy
of chemotherapy have been used to treat patients with medulloblastoma rather than classic or desmoplastic residual or recurrent medulloblastoma or with primitive medulloblastomas. Preliminary data suggest that large-cell neuroectodermal tumours (PNET). Single-agent melphalan anaplastic tumours have other features (eg, overexpression used in tandem with bone-marrow transplantation for of the ERBB2 receptor and c-MYC amplifi cation) that children (median age 4·1 years) diagnosed with supra- tentorial PNET showed response in 11 of 14 assessable Abnormalities in the WNT and SHH pathways are patients. These newly diagnosed patients were not given important in the pathogenesis of medulloblastoma.31 We radiotherapy before high-dose chemotherapy.28 French have previously shown that medulloblastoma consists of investigators have also used high-dose chemotherapy with several distinct molecular subgroups.16 In this study, no busulfan and thiotepa as a salvage regimen for infants who tumour showed evidence of activation of both the WNT relapse on standard chemotherapy regimens. Preliminary and SHH pathways, which lends support to our previous results of this approach showed that this strategy could be report16 that these alterations are mutually exclusive in useful in salvaging infants with local relapse if used in medulloblastoma.
conjunction with local radiotherapy to the tumour bed. Notably, we showed that tumours containing mutations Patients with metastatic disease did not seem to benefi t in CTNNB1 express a unique gene expression signature; from this approach.29 have large portions of chromosome 6 deleted; and tend to Despite reports of 5–10% mortality due to the toxic eff ects be found in older children. Patients with tumours of high-dose chemotherapy used as a salvage regimen for harbouring mutations in CTNNB1 have also been reported children with recurrent brain tumours,6–10,28,29 none of our to show an improved prognosis,32 which our data confi rm. patients died from regimen-related toxic eff ects. The fact Together, these fi ndings support the hypothesis that that 92% of our patients completed all four cycles of high- medulloblastoma containing activating mutations in dose chemotherapy suggests that our approach is safe, CTNNB1 represent a molecularly and clinically distinct feasible to deliver, and improves outcome for patients with disease subset with a favourable clinical outcome. Further studies could pilot reduction in treatment in patients with Several investigators have shown that patients with large- cell anaplastic medulloblastoma have low event-free We have shown that more than 70% of high-risk and survival. Our data confi rm this fi nding but also show that average-risk medulloblastomas are curable. However, long-for average-risk patients, the risk of treatment failure is term treatment-related morbidity, especially neurocognitive almost quadrupled if they have large-cell anaplastic and functional defi cit, remains a major hurdle that impairs http://oncology.thelancet.com Vol 7 October 2006
Articles
patients’ ability to complete their education and gain 14 Strother D, Ashley D, Kellie SJ, et al. Feasibility of four consecutive fi nancial and social independence.33,34 Patients could be high-dose chemotherapy cycles with stem-cell rescue for patients with The Lancet Oncology, please
newly diagnosed medulloblastoma or supratentorial primitive cured with decreased morbidity by the use of reduced neuroectodermal tumor after craniospinal radiotherapy: results of a treatment in patients with clinically and molecularly collaborative trial. J Clin Oncol 2001; 19: 2696–704.
favourable disease and by use of drugs that target the 15 Chang C, Housepain E, Herbert CJ. An operative staging system and a megavoltage radiotherapeutic technique for cerebellar genetic alterations that cause medulloblastoma.
medulloblastoma. Radiology 1969; 93: 1351–59.
Confl icts of interest
16 Thompson MC, Fuller C, Hogg T, et al. Genomics identifi es We declare no confl icts of interest.
medulloblastoma subgroups that are enriched for specifi c genetic
alterations. J Clin Oncol 2006; 24: 1924–31.
Contributors
17 Stewart CF, Iacono LC, Chintagumpala M, et al. Results of a phase II A Gajjar was principal investigator of the study; and participated in the upfront window of pharmacokinetically guided topotecan in high-risk provision of study materials or patients, collection and assembly of data, medulloblastoma and supratentorial primitive neuroectodermal data analysis and interpretation, manuscript writing, and fi nal approval of tumor. J Clin Oncol 2004; 22: 3357–65.
manuscript. M Chintagumpala, D Ashley, S Kellie, L E Kun, T E Merchant,
18 Kleihues P, Cavanee WK. World Health Organization classifi cation of S Woo, G Wheeler, V Ahem, M J Krasin, M Fouladi, A Broniscer, tumours: pathology and genetics of tumours of the nervous system. R Krance, G A Hale, C F Stewart, R Dauser, R A Sanford, C Fuller, C Lau, and R J Gilbertson participated in the provision of study materials and 19 National Cancer Institute. Common toxicity criteria v2.0, April 30, 1999. patients, collection and assembly of data, data analysis and interpretation, http://ctep.cancer.gov/reporting/ctc.html (accessed Aug 26, 2006).
manuscript writing, and fi nal approval of manuscript. J M Boyett and 20 Heideman RL, Kovnar EH, Kellie SJ, et al. Preirradiation D Wallace participated in the collection and assembly of data, data analysis chemotherapy with Carboplatin and Etoposide in newly diagnosed and interpretation, manuscript writing, and fi nal approval of manuscript.
embryonal pediatric CNS tumors. J Clin Oncol 1995; 13: 2247–54.
21 Peto R, Pike MC, Armitage P, et al. Design and analysis of Acknowledgments
randomized clinical trials requiring prolonged observation of each This work was supported by a Cancer Center (CORE) Support Grant (CA patient. II: analyses and examples. Br J Cancer 1977; 35: 1–39.
21765) from the National Institutes of Health, Musicians against 22 Packer RJ, Goldwein J, Nicholson HS, et al. Treatment of children Childhood Cancer (MACC), the Noyes Foundation, and the American with medulloblastoma with reduced-dose craniospinal radiation Lebanese Syrian Associated Charities (ALSAC).
therapy and adjuvant chemotherapy. A Children’s Cancer Group
Study. J Clin Oncol 1999; 17: 2127–36.
References
1
Hughes EN, Shilito J, Sallan SE, et al. Medulloblastoma at the Joint 23 Packer RJ, Gajjar A, Vezina G, et al. Phase III prospective randomized Center for Radiation Therapy between 1968 and 1984: the infl uence of study of craniospinal radiation therapy followed by one of two radiation dose and pattern of failure. Cancer 1988; 61: 1992–98.
adjuvant chemotherapy regimens for newly-diagnosed average risk
medulloblastoma. J Clin Oncol 2006; 24: 4202–08.
Gottardo NG, Gajjar A. Current therapy for medulloblastoma.
Curr Treat Options Neurol 2006, 8: 319–34.
24 Miralbell R, Fitzgerald TJ, Laurie F, et al. Radiotherapy in pediatric medulloblastoma: quality assessment of pediatric oncology group trial Zeltzer PM, Boyett JM, Finlay JL, et al. Metastasis stage, adjuvant 9031. Int J Radiat Oncol Biol Phys 2006; 64: 1325–30.
treatment, and residual tumor are prognostic factors for medulloblastoma in children: conclusions from the Children’s Cancer 25 Oyharcabal-Bourden V, Khalifa C, Gentet JC, et al. Standard-risk Group 921 randomized phase III study. J Clin Oncol 1999; 17: 832–45.
medulloblastoma treated with adjuvant chemotherapy followed by reduced-dose craniospinal radiation therapy: a French Society of Taylor RE, Bailey CC, Robinson KJ, et al. Outcome of patients with Pediatric Oncology Study. J Clin Oncol 2005; 23: 4726–34.
metastatic (M2-3) medulloblastoma treated with SIOP/UKCCSG
PNET-3 chemotherapy. Eur J Cancer 2005; 41: 727–34.
26 Taylor RE, Bailey CC, Robinson KK, et al. Results of a randomized study of pre-irradiation chemotherapy vs radiotherapy regimen alone Kortmann RD, Kuhl J, Timmermann LA, et al. Postoperative for non-metastatic (M –M ) medulloblastoma. The SIOP/ UKCCSG neoadjuvant chemotherapy before radiotherapy as compared to PNET-3 study. J Clin Oncol 2003; 21: 1581–91.
immediate radiotherapy followed by maintenance chemotherapy in the treatment of medulloblastoma in childhood: Results of the 27 Packer RJ, Sutton LN, Elterman R, et al. Outcome for children with German prospective randomized trial HIT 91. Int J Radiat Oncol Biol medulloblastoma treated with radiation and cisplatin, CCNU and Phys 2000; 46: 269–79.
vincristine chemotherapy. J Neurosurg 1994, 81: 690–98.
Finaly JL, Goldman S, Wong MC, et al. Pilot study of high-dose 28 Vassal G, Tranchand B, Valteau-Couanet D, et al. Pharmacodynamics thiotepa and etoposide with autologous bone marrow rescue in of tandem high-dose melphalan with peripheral blood stem cell children and young adults with recurrent CNS tumors: the Children’s transplantation in children with Neuroblastoma and Cancer Study Group. J Clin Oncol 1996; 14: 2495–503.
medulloblastoma. Bone Marrow Transplant 2001; 27: 471–77.
Graham ML, Herndon JE II, Casey JR, et al. High-dose chemotherapy 29 Dupuis-Girod S, Hartmann O, Benhamou E, et al. Will high dose with autologous stem-cell rescue in patients with recurrent and high- chemotherapy followed by autologous bone marrow transplantation risk pediatric brain tumors. J Clin Oncol 1997; 15: 1814–23.
supplant cranio-spinal irradiation in young children treated for
medulloblastoma? J Neuroonocol 1996; 27: 87–98.
Dunkel I, Boyett JM, Yates A, et al. High-dose carboplatin, thiotepa and etoposide with autologous stem-cell rescue for patients with 30 McManamy CS, Lamont JM, Taylor RE, et al. Morphophenotypic recurrent medulloblastoma: Children’s Cancer Group. J Clin Oncol variation predicts clinical behavior in childhood non-desmoplastic 1998; 16: 222–28.
medulloblastoma. J Neuropathol Exp Neurol 2003; 62: 627–32.
Gururangan S, Dunkel IJ, Goldman S, et al. Myeloablative 31 Gilbertson RJ. Medulloblastoma: signalling a change in treatment. chemotherapy with autologous bone marrow rescue in young Lancet Oncol 2004; 5: 209–18.
children with recurrent brain tumors. J Clin Oncol 1998; 16: 2486–93.
32 Ellison DW, Onilude OE, Lindsey JC, et al. β catenin status predicts a 10 Mahoney D, Strother D, Camitta B, et al. High-dose melphalan and favorable outcome in childhood medulloblastoma: the United cyclophosphamide with autologous bone marrow rescue for recurrent Kingdom Children’s Cancer Study Group Brain Tumor Committee. progressive malignant brain tumors in children. A pilot Pediatric J Clin Oncol 2005; 23: 7951–57.
Oncology Group study. J Clin Oncol 1996; 14: 382–88.
33 Mulhern RK, Merchant TE, Gajjar A, et al. Late neurocognitive Abrahemsen TG, Lange BJ, Packer RJ, et al. A phase I and II trial of sequelae in survivors of brain tumours in childhood. Lancet Oncol dose-intensifi ed cyclophosphamide and GM-CSF in pediatric 2004; 5: 399–408.
malignant brain tumors. J Pediatr Hematol Oncol 1995; 17: 134–39.
34 Fouladi M, Chintagumpala M, Laningham FH, et al. White matter 12 Lachance DH, Oette D, Schold SC Jr, et al. Dose escalation trial of lesions detected by magnetic resonance imaging after radiotherapy cyclophosphamide with saragamostim in the treatment of central and high-dose chemotherapy in children with medulloblastoma or nervous system (CNS) neoplasms. Med Pediatr Oncol 1995; 24: 241–47.
primitive neuroectodermal tumor. J Clin Oncol 2004; 22: 4551–60.
13 Frei E III, Canellos GP. Dose a critical factor in cancer chemotherapy. Am J Med 1980; 69: 585–94.
http://oncology.thelancet.com Vol 7 October 2006

Source: http://neurooncologia.ru/sites/default/files/SJMB-96_results.pdf