Given the rarity of ATRTs and the variety of treatment regimens used to date, no standard therapeutic approach has been established. Published case series often include patients treated with multiple therapeutic approaches, making standardization of therapy difficult. Table summarizes some of the larger clinical trials that included ATRT patients or that were designed specifically for ATRT. Early complexity determining the best therapeutic approach was evident in the reported proceedings of the tumor board of The Children’s Hospital of Philadelphia (CHOP) in 1991 (Perilongo et al., 1991
). The group described a 5-year-old patient with a CNS primary rhabdoid tumor and detailed their consideration of therapeutic approaches including review of children with rhabdoid tumors of the kidney. Ultimately, they opted to treat with 3600 cGy craniospinal radiation with an unspecified boost dose to the primary site along with chemotherapy that was currently in use on the infant brain tumor therapeutic trials, including cisplatin, cyclophosphamide, vincristine, and etoposide.
Published reviews and clinical trials that included or were specific to atypical teratoid rhabdoid tumors.
In response to the report from CHOP, Weinblatt and Kochen (1992
) from Cornell submitted a letter describing a patient they treated in 1985 with a primary CNS rhabdoid tumor with gross total resection (GTR), 4140 cGy focal radiation, and intensive chemotherapy as per the Intergroup Rhabdomyosarcoma III (IRS-III) therapy, including weekly vincristine during radiation, actinomycin-D, doxorubicin, and triple intrathecal chemotherapy with hydrocortisone, methotrexate, and cytosine arabinoside. An additional three cases successfully treated with IRS-III were then reported in 1995 by Ohio State University (Olson et al., 1995
). This approach was justified because ATRT was thought to be similar to parameningeal rhabdomyosarcomas, requiring more aggressive therapy, and regimen 36 was chosen because it was intensive chemotherapy that was easily adaptable to radiation and triple intrathecal chemotherapy. This group also summarized the 18 cases of primary CNS MRTs previously reported in the literature to date, showing the varied approaches to therapy, and the need for a more standardized approach.
Dana-Farber Cancer Institute (DFCI) decided to return to the original early reports of success with IRS-III-based regimens to treat two new ATRT patients and two with recurrent disease between December 1999 and April 2002 due to unpublished reports of failures with modifications from the original therapy (Zimmerman et al., 2005
). DFCI modifications included focal stereotactic RT without a craniospinal dose for patients less than 3
years old, substitution of dacarbazine with temozolomide, and addition of dexrazoxane for cardioprotection in higher cumulative doses of doxorubicin. All four patients were alive at the time of reporting at a median 44.5
months after diagnosis and a median 26.5
months after recurrence. A later update from the group at DFCI stated that three of the four were alive at a median 6.5
years after completion of therapy. DFCI then proceeded with a phase II study between February 2004 and September 2006 with a modified IRS-III protocol and treated 20 patients with ATRT (Chi et al., 2009
). Eight of the 20 patients had relapses by the time of publication, giving a 1-year progression-free survival (PFS) rate of 70
10% and OS of 75
10% and 2-year PFS of 53
13% and OS 70%
10%. Univariate analysis showed that PFS and OS were significantly influenced by the extent of resection. OS was also affected by tumor location, and patients with posterior fossa tumors had better survival. The reported PFS and OS were significantly better than those seen in other clinical trials but, due to small numbers, it was impossible to make comparisons to determine why there was such an improvement. This report did, however, point to improved survival with intensified chemotherapy that included intrathecal administration along with focal radiation for those patients younger than 3
years old. It would also have been useful with larger numbers to separate the overall survival data based upon age, as others have shown that survival is improved for older patients.
Modified IRS-III therapies include intrathecal chemotherapy as well as multiagent chemotherapy and focal radiation in patients who have non-metastatic disease. Intrathecal chemotherapy may have potential benefit as an additional means to avoid radiation or to intensify therapy in patients who are not candidates for craniospinal radiation. A meta-analysis by Athale et al. (2009
) showed that even without GTR, patients who received multiagent chemotherapy fared better, but this effect was most prominent in those less than 3
years old who did not get radiation. Without radiation, intrathecal chemotherapy also made a significant difference in overall survival (OS 10.5
months versus 6.5
An additional approach to early therapy of ATRT was explored as these patients were included in national infant brain tumor clinical trials. The North American Children’s Cancer Group from April 1993 through June 1997 enrolled 299 children less than 3
years old with multiple tumor types on protocol CCG9921 (Geyer et al., 2005
). This regimen included two induction courses with ifosfamide or cyclophosphamide along with vincristine, cisplatin, and etoposide. Induction was followed by maintenance with vincristine, etoposide, carboplatin, and cyclophosphamide. The study included 28 rhabdoid tumors (9.4% of patients enrolled), and 24 of those had treatment failures. The 1-year and 5-year EFS rates were 32
9% and 14
7%, respectively, and the 5-year OS was 29
9%. Interpretation of factors associated with prognosis was difficult due to the small numbers in this group. During the same period, the Pediatric Oncology Group was investigating the use of standard versus dose-intensified chemotherapy to delay radiation in young children on POG 9233/34 (Baby POG 2). The study enrolled 36 patients with ATRT. Chemotherapy included cyclophosphamide, vincristine, cisplatin, and etoposide. Patients on the dose-intensified arm had better responses, but all patients with rhabdoid tumors ultimately died, with a median survival of 6.7
months (personal communication, Douglas Strother).
Researchers from St. Jude Children’s Research Hospital in 2005 reported a retrospective review of 31 patients with ATRT treated between 1984 and 2003 (Tekautz et al., 2005
). As expected, the patients were treated with multiple different approaches over the 20-year period, but overall they determined that outcomes were better for patients who were older than 3
years at diagnosis and those who had received craniospinal radiation and high-dose alkylator-based chemotherapy up front with a 2-year OS of 89
11%. Five of the nine patients older than 3
years were alive without recurrence at the time of publication at a median of 2.2
years, and it is important to note that seven of those patients had craniospinal radiation. In contrast, only four of 22 patients less than three were without relapse. One of the four died of a surgical complication; of the remaining three, two received RT.
The Canadian Pediatric Brain Tumor Consortium recently published a retrospective review of patients with ATRT treated from 1995 through 2007 (Lafay-Cousin et al., 2012
). They identified 50 patients, and although there were multiple different therapies during the period, they were able to make several conclusions from their analysis. As expected, the prognosis was much worse for patients younger than 12
months. Patients with GTR had better survival at 2
years at 60
12.6% versus 21.7
0.03). They did not find a significant difference in survival related to radiation, but they did find that patients treated with high-dose chemotherapy (HDC) had better 2-year OS at 47.9
12.1% versus 27.3
There seems to be improved survival for those patients treated on IRS-III based de novo
therapy and high-dose alkylating agent compared to other chemotherapy approaches although numbers are small in all series and it is difficult to separate the role the modifications of the IRS-III regimen. As stated earlier, intrathecal chemotherapy has been associated with improved survival in those patients that did not receive radiation. In the IRS-III based regimen, the combination of intrathecal chemotherapy with focal radiation in those patients less than 3
years of age may be an effective strategy for disease control, but intrathecal chemotherapy in those patients older than 3
years may not be needed since craniospinal radiation is an acceptable option. It is clear that intensive systemic chemotherapy alone as a method to avoid radiation is not effective in ATRT and radiation must be considered much earlier in therapy than previously thought appropriate, but with continued respect for long term effects.
Since 2008, the Children’s Oncology Group (COG) has been enrolling patients with ATRT on a clinical trial that incorporates induction chemotherapy with high-dose methotrexate, focal radiation in patients as young as 6
months for infratentorial tumors and 12
months for supratentorial tumors, and three cycles of consolidation with thiotepa and carboplatin with autologous stem cell support. This therapeutic protocol thus incorporates focal radiation in younger patients than on previous COG protocols. High-dose methotrexate was included based on data from the Head Start (HS) protocols discussed later and the three cycles of consolidation are based on CCG 99703 for which data has yet to be published. The study was closed for 1
year due to the toxic death of one patient but has been reopened with amendments concerning pulmonary toxicity monitoring and increased time between consolidation courses. Enrollment is ongoing, and the accrual rate is as expected for this rare tumor.