SEARCH

SEARCH BY CITATION

Keywords:

  • medulloblastoma;
  • children;
  • twinning;
  • low-income countries

Abstract

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Successful twinning initiatives have been reported in childhood leukemia. Pediatric neuro-oncology requires a complex multidisciplinary approach and the feasibility of similar twinning programs is unknown. Twinning between King Hussein Cancer Center in Amman and the Hospital for Sick Children in Toronto started with e-mail communications, and subsequently included monthly videoconferences and exchanges between institutions. The outcome of 35 newly diagnosed medulloblastoma patients (22 high-risk and 13 average-risk) treated during this period is reported. The 3-year overall survival for average risk and high-risk patients was 100 and 81%, respectively. This experience suggests that twinning may facilitate the implementation of multidisciplinary neuro-oncology programs in low-income countries. Videoconferencing allows interactive exchanges with a significant learning impact. © 2007 Wiley-Liss, Inc.

More than 80% of children live in low-income countries, where the main health care priority is to address the problems of malnutrition, infection and parasitic diseases. Children with cancer in these countries have no or little access to effective therapy, and the geographic inequality of pediatric oncology around the world has triggered the development of twinning programs between institutions in high- and low-income countries.1–3 Significant reduction in mortality rates have been achieved through twinning in childhood lymphoblastic leukaemias, and ongoing experiences are in place in South America, in Africa or in Asia.4–6

Incidence and survival rates in children with brain tumors in many low-income countries are largely speculative. Many cases are undiagnosed and access to treatment is restricted by availability of adequate health structures, although most countries with limited resources have neurosurgical facilities, radiation equipments and pediatric oncology units with experience in delivering chemotherapy. The main challenge for the development of twinning experiences in pediatric brain tumors is the complexity of the care required for these patients. The benefit of a twinning program is measured by its impact on survival. The twinning program between the King Hussein Cancer Center in Amman (KHCC) and the Hospital for Sick Children in Toronto was initiated in 2003,7 and the current report focuses on outcomes for the medulloblastoma population. The study was approved by the Research Ethics Board of KHCC.

Material and methods

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Institutions and context

The Hospital for Sick Children was originally built in 1875. It is a tertiary and quaternary care pediatric academic health sciences center with a comprehensive pediatric brain tumor program, which includes 5 pediatric neurosurgeons, 4 pediatric neuro-oncologists and 3 radiation oncologists. King Hussein Cancer Center, founded in 1997, is a nongovernmental, not-for-profit comprehensive 148 beds oncology center, with a pediatric oncology unit of 29 beds with more than 200 new patients per year. The Pediatric brain tumor program includes 1 pediatric oncologist, 1 part time pediatric neurosurgeon and 1 radiation oncologist. The hospital serves a large population, with 75% of patients coming from Jordan and 25% from other countries of the Middle East. Jordan is ranked 132nd nation based on the GDP per capita8 with a population of 5.9 million and over 33% of the population under 15-year-old. Prior to 2002, there was no dedicated pediatric neuro-oncology program in Jordan. Few pediatric brain tumor patients were seen at KHCC, and a large majority of children were treated in private institutions and other government hospitals without pediatric neuro-oncology expertise. As there is no national registry and no data available in the private sector, information on treatment and outcome is only speculative. Metastatic work up (MRI scan of the spine and CSF examination) was infrequently performed. When postoperative treatment was offered, most children were treated with radiation only, using either focal radiation to the posterior fossa or craniospinal radiation. The neuro-oncology program started in December 2002 with the appointment of a pediatric oncologist (IQ) dedicated to the brain tumor and retinoblastoma population.

Twinning

The twinning experience started in 2003 with electronic e-mails. Between 2003 and 2005 there were 356 e-mails sent from KHCC to the Hospital for Sick Children. In addition, there were a total of 5 weeks exchange visits at either institution since starting cooperation. The twinning was facilitated by the initiation of a videoconference activity in December 2004. Between December 2004 and December 2006, 26 one hour videoconferencing sessions were held.

Management of medulloblastoma patients

The choice of the protocols was decided after a careful review of the literature and a thorough assessment of the resources available in our center. For high-risk patients the consensus was to consider craniospinal radiotherapy (CSI), followed by 3 cycles of cisplatin and etoposide and 8 cycles of cyclophosphamide and vincristine according to a previous protocol of the Pediatric Oncology Group.9

For average-risk medulloblastoma patients, the decision was to use reduced dose CSI with concomitant weekly vincristine followed by 8 courses of cisplatin, lomustine and vincristine following the promising results of a pilot study conducted by the Children Oncology Group in children less than 10 years of age.10 Laboratory work (blood counts, electrolytes) was performed every other week due to logistical and financial issues. Because of the limited experience of central lines, chemotherapy was administered through peripheral lines when ever possible. Only 1 patient required central line insertion.

For radiation, a linear accelerator (Philips SL-25 and Elekta precise, Netherlands) was used. The dose of CSI was 23.4 Gy in 13 fractions for average-risk patients and 36–39.6 Gy in 20–22 fractions for high-risk patients. The total dose to the posterior fossa was 54 Gy in 30 fractions. On admission, all patients had complete metastatic work up (most often not done or incomplete prior to admission) and risk groups were defined according to the North American standards.8, 9

Results

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Population

Between December 2002 and December 2006, 49 medulloblastoma patients (10 nonJordanians) were seen at our institution. We included in this report all patients who were treated in first instance with CSI, regardless of their age. Six newly diagnosed infants (all Jordanian) treated with postoperative chemotherapy only, and 5 patients with recurrent medulloblastoma were excluded. Three patients whose families declined the treatment recommended were also excluded. Overall, a total of 35 (7 nonJordanians) newly diagnosed medulloblastoma patients are included in this report. Patients' characteristics are detailed in Table I. When patients were referred from other institutions, the pathology was reviewed by our neuropathologist (MAH). Twenty-four patients had a classical histology, 10 desmoplastic and 1 had anaplastic large cell medulloblastoma. All but 2 patients were older than 36 months of age at the time of diagnosis and median age was 68 months (range: 31–223 months). All surviving patients were seen at KHCC or contacted during the spring of 2007.

Table I. Patients' Characteristics
 High riskAverage risk
  1. Bx, biopsy (<50%); STR, subtotal resection (>50% and 95%<); NTR, near total resection (>95%); GTR, gross total resection; DOT, dead of toxicity; DOD, dead of disease.

Number of patients2213
Median age (range)71 (35–223)60 (31–164)
Initial surgery at KHCC85
Resection  
 GTR712
 NTR11
 STR10 
 Bx4 
Metastatic stage  
 M0713
 M12 
 M2/313 
Delay surgery-radiation median (range)49 (17–144)45 (27–117)
 Initial surgery at KHCC43 (17–58)34 (27–55)
 Initial surgery outside KHCC53 (27–144)59 (35–117)
Major changes in chemotherapy  
 Refusal of chemotherapy20
 Dose adjustments54
 Early discontinuation of chemotherapy06
Outcome [median follow-up in months (range)]27 (4–55)25 (5–48)
 Alive without progression1913
 DOD1 
 DOT1 
 Lost to follow-up1 
 Patients discussed at videoconference33

Twenty-two patients underwent surgical resection prior to their admission at KHCC. Nineteen patients underwent ventriculoperitoneal shunt placement. Fifteen (79%) of these shunts were inserted prior to referral to our institution. A significant difference was observed in the time between surgery and initiation of radiation between patients operated outside and at our institution (p < 0.05, Fig. 1).

thumbnail image

Figure 1. Delay between surgery and radiation according to the place of surgery.

Download figure to PowerPoint

Twenty-two patients were treated as high-risk medulloblastoma. Seven cases were staged as M0, 2 M1 and 13 M3. Two patients originally stratified as standard risk were eventually upgraded to high-risk because of refusal of chemotherapy, and evidence of local progression on repeat imaging while waiting to start radiotherapy. At a median follow-up of 27 months (range: 2–52 months), 19 patients are alive without progression. Two patients died, 1 from disease progression and the second patient from radiation necrosis. One patient is lost to follow-up.

Thirteen patients (37%) were treated as standard risk medulloblastoma. Two Palestinian patients continued their chemotherapy in West Bank and were reported to have completed all 8 cycles of chemotherapy. A third Palestinian child who was supposed to receive his chemotherapy in the West Bank did not receive any further treatment. One patient received only one cycle of chemotherapy and developed prolonged severe thrombocytopenia, which precluded further administration of the protocol. With a median follow-up of 25 months (range: 5–48 months), there has been no relapse and no death reported in this group and no patient was lost to follow-up. The 3-year overall survival for average-risk and high-risk patients was 100 and 81% (CI: 57–99%, Fig. 2).

thumbnail image

Figure 2. Survival estimates of average risk and high-risk patients.

Download figure to PowerPoint

Toxicity

Because of the intensity of chemotherapy and the absence of satellite centers except for the West Bank, families were advised to stay in town between courses. When families had no relatives locally, KHCC Foundation was covering the cost of accommodation in a nearby hotel. Out of 261 cycles of chemotherapy delivered, there were 76 nonchemotherapy related admissions (including 39 febrile neutropenic episodes) in 23 patients. Thirteen patients required at least one red blood cell transfusion (range 1–6) and 2 patients received platelets transfusions. Thirty-two admissions (42%) occurred during the first year of the neurooncology program. Most dose modifications (Table I) occurred in average-risk patients: in the high-risk group, 3 patients had dose reduction or discontinuation of vincristine, the dose of cyclophosphamide was reduced in another patient, and in 1 patient cisplatin was changed to carboplatin because of renal insufficiency. In the average-risk group, a total of 13 cycles were omitted in 4 patients because of severe toxicities. One patient developed prolonged thrombocytopenia after only one cycle of chemotherapy and subsequently received 20 weeks of vincristine. The dose of cisplatin was altered in 18 cycles (9 reductions, 9 discontinuations) because of ototoxicity (6 patients) or kidney toxicity (1 patient). The dose of vincristine dose was modified in 2 patients (1 for ileus and 1 for neuropathy).

Impact of videoconference

Between December 2004 and December 2006, 26 sessions of videoconferencing were held. One hundred two cases (3.9 cases per session) and 85 patients were discussed (11 patients were discussed twice, 1 patient discussed three times, and 1 patient discussed 5 times). Out of 85 patients, 11 (13%) had medulloblastoma, including 6 newly diagnosed patients, 2 patients on follow-up and 3 patients with recurrence. In 1 patient with suspected isolated cervical relapse, the hypothesis of radiation necrosis secondary to an overlap of the fields was suggested. This prompted a comprehensive review of the radiation fields and a 6-cm overlap was confirmed. This experience had a major impact on the radiotherapy practice at KHCC. Other major changes included recommendation of repeat surgery after partial resection in 1 case, avoidance of radiation in an infant with desmoplastic medulloblastoma, a trial of oral chemotherapy prior to radiation in a child with extensive metastatic disease and poor general condition at diagnosis, and upfront CSI (rather than chemotherapy and delayed radiation) in a child with significant delays due to repeated CSF infections. The videoconference also allowed interactive discussions between the members of the teams, in particular technical discussions concerning surgical approaches and indication for CSF diversion, the importance of maximal resection and the role of second look surgery in nonmetastatic medulloblastoma, the evaluation of metastatic disease on MRI and the difference between nonspecific abnormalities at the spinal level (such as clumping of the roots or vascular enhancement) and nodular or leptomeningeal metastatic disease.

Discussion

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Survival in medulloblastoma has dramatically increased with the introduction of postoperative CSI11 and the development of cooperative protocols using multi-agent chemotherapy.9, 10, 12, 13 The best survival rates at 5 years in average-risk and high-risk medulloblastoma patients are currently above 80 and 70%, respectively.9, 10, 12, 13 However, these data relate to studies conducted in Western countries involving selected patients enrolled based on strict eligibility criteria. Survival rates in low-income countries are only speculative, as there has been no prospective study conducted in the developing world in any type of pediatric brain tumor. There have been some publications on pediatric brain tumors in developing countries.14, 15 However, incidence in these series is far below the expected numbers based on corresponding demographic figures and this suggests referral bias or underreporting. The reported annual average incidence of medulloblastoma is 3–6 per million in children and young adults less 20 years of age.16 With 34 Jordanian children with medulloblastoma seen at KHCC during the 3.5-year study period, our patient population correlates with the expected number of cases diagnosed in Jordan over the period of time. Late diagnoses and significant delays between initial surgery and referral to our institution may explain the large proportion of metastatic patients in this cohort (15/35 M+, 42%). However, the proportion of high-risk patient decreased over time with a ratio high risk/average risk of 9/1 for the period 2002–2003 versus 13/12 for 2004–2006, and this trend may reflect the impact of the development of the neuro-oncology program on clinical practice regionally and nationally. Other changes were observed over time, in particular a gradual decrease in the time between initial surgery and initiation of radiation (Fig. 3).

thumbnail image

Figure 3. Delay surgery-radiation. Evolution over time.

Download figure to PowerPoint

The role of the twinning initiative in this development has been valuable. The implementation of appropriate average- and high-risk protocols was extensively discussed with regard of the resources available and the supportive care needed. Identification of technical problems concerning radiation techniques allowed an extensive review of the service provided. Ongoing exchanges during the teleconferences contributed to refine our approach to staging and interpretation of imaging studies. This opportunity of reviewing MRI studies in a real time manner suggests that telemedicine may also potentially benefit prospective clinical trials. In a recent prospective randomized trial, central radiology review revealed that 10% of medulloblastoma patients enrolled were ineligible,12 and outcome of inappropriately staged patients was significantly worse (5 year event free survival 36 vs. 83%) compared with appropriately staged patients. Introduction of real time review using telemedicine may avoid inappropriate treatment allocation in a disease, where there is “no room for complacency.”17 The videoconference was also an opportunity to discuss individual situations, when protocols were deemed to be inappropriate for patients in poor condition or with specific issues. These patients are usually noneligible for prospective clinical trials and there are no guidelines for these complex situations. Sharing and discussing these experiences has been a unique learning experience for both teams.

Many other challenges can affect the development of new programs in countries with limited resources. Hierarchical health care systems may see the introduction of new techniques by young and Western trained physicians as a potential threat. The support of the executive board of the Jordanian institution in this experience has been critical. Building a multidisciplinary team is challenging in a Health Care system, which is divided into sectors with limited cross communication. The example of a cohesive multidisciplinary team (at the Hospital for Sick Children) has been a driving inspiration for the members of the new neuro-oncology program at KHCC, and communication between Jordanian pediatric oncologists, neuroradiologists, pathologists, neurosurgeons and radiation oncologist has greatly benefited from this twinning initiative.

The follow-up of this cohort is relatively short and survival data cannot be compared to the results of recent reports of prospective studies with longer follow-up.9, 10, 12, 13 However, specific aspects of the development of this program are of critical interest for pediatric oncology units in low-income countries who may consider similar initiatives. Early results observed in this unselected cohort of patients are promising and suggest that outside leukemia and lymphoma, the survival gap between high-income and low-income countries in pediatric oncology can also be substantially reduced through twinning initiatives.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

The authors acknowledge the current and previous Directors Generals for King Hussein Cancer Center, Dr. Mahmoud Sarhan and Dr. Samir Khleif (KHCC director general, from November 2002 till February 2006 and currently with the NIH) for their vision and support for this initiative, the Information Technical Staff at King Hussein Cancer Center (especially Ameen Harb and Shamil Ansouqa) and at the Hospital for Sick Children (Agnes Cheng-Tsallis and Joanne Reid) for their technical support in the videoconferencing program. From the Hospital for Sick Children, we acknowledge Dr. Abhaya Kulkarni and Dr. Peter Dirks from the Division of Neurosurgery division and from Princess Margaret Hospital, Dr. Normand Laperriere, radiation oncologist, for his valuable recommendations. Finally, the authors acknowledge Ms. Tharwa Bilbeisi's invaluable contribution for editing the manuscript.

References

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References