The manuscript by Jubran et al brings to light the many controversies surrounding the management of intracranial germ cell tumors. Perhaps the most controversial issues are the roles of radiotherapy and chemotherapy in the treatment of pure germinoma and nongerminomatous germ cell tumor (NGGCT). Radiotherapy Alone in Pure Germinoma
As discussed by Jubran et al, treatment of pure germinoma with radiotherapy (RT) alone has a long track record of survival exceeding 90%.[1,2] Traditionally, craniospinal RT has been used; however, more recent data do not support treating the spine in this setting.[2-4] Table 1 shows a compilation of studies predominantly treated with RT alone.[1,2,4-10] Relapse rates are low when treating with craniospinal and whole-brain RT. The data also suggest that treating with wholeventricular RT does not increase the number of relapses, although the number of patients treated with this RT field is low and predominantly from one study.[9] Treatment with involvedfield RT (primary site with a margin) resulted in a 25.9% relapse rate. Based on the above data, the whole ventricle should be the minimum volume treated when using RT alone. As noted, however, the experience is scant and treatment with this volume should be approached with caution. The next Children's Oncology Group (COG) trial will use whole-ventricle RT followed by a primary site boost as the radiotherapy- alone arm and will compare this to the use of neoadjuvant chemotherapy followed by primary site RT. Appropriate Dose
As to the appropriate dose, numerous studies have shown that that doses ≥ 50 Gy to the primary site provide excellent local control.[3,4] Doses < 40 Gy produced a 52% brain failure rate in the Mayo Clinic study and should not be used when treating with RT alone. As mentioned by Jubran etal, doses of 40 to 45 Gy have provided excellent control in more recent studies and should be the new standard.[ 1,11,12] With regard to prophylactic doses to uninvolved sites such as the craniospinal axis, whole-brain, or whole-ventricular system, the article mentioned that doses can be lowered to 30 Gy without detriment. More recent studies have shown that prophylactic doses can be further lowered. Hardenbergh et al found no failures in nine patients receiving wholebrain doses of 19.8 to 25.5 Gy.[8] At St. Jude Children's Research Hospital, Merchant showed no local failures with a median craniospinal dose of 25.6 Gy.[10] Maity et al reported no failures in five patients receiving doses from 18 to 19.8 Gy and six patients treated with doses from 23.4 to 27 Gy. Based on these results, one can argue for a lower prophylactic dose of approximately 25 Gy. Chemotherapy Options
Concerns regarding the long-term toxicity of radiotherapy have led investigators to search for treatments that would eliminate or minimize RT. Most papers regarding the detrimental effects of RT on neurocognitive and hormonal function have dealt with primary brain tumors in young children, using higher doses of RT. The effects of RT in children with germinomas may be less severe because of different host and treatment factors. Children with germinomas tend to be in the adolescent age group, unlike children with other primary brain tumors such as medulloblastoma and ependymoma. Moreover, radiotherapy doses employed in germinoma are lower compared to those used in other brain tumors. That said, many children with germinoma have endocrine problems even before starting any treatment.[1,8,10] Two recent studies show excellent late outcomes after craniospinal RT alone for intracranial germinoma. At the Children's Hospital of Philadelphia, 22 patients were evaluated for quality of life and were found to have higher mental composite scale scores compared to controls. All of the patients were in high school or had completed high school; nine were in college or completed college and five had advanced degrees. Patients were normally proportioned for height and weight, although the women tended to be shorter.[13] Merchant et al found no significant difference in pre- and post- RT full-scale, verbal and performance IQ scores and pre- and post-RT statures and weights among 12 children treated with craniospinal RT.[10] The First and Second International Central Nervous System Germ Cell Tumor Studies found 41% to 47% event-free survivals using a primary chemotherapy approach.[14,15] The first study used four cycles of carboplatin, etoposide, and bleomycin followed by two cycles of the same drugs with or without cyclophosphamide intensification. The second trial used more intensive cisplatin- and cyclophosphamide- based chemotherapy. What is concerning is the toxicity of treatments in these two studies: In the first study, 7 of 71 patients died of chemotherapy- related toxicity, whereas in the second study, 3 of 19 patients died of treatment-related complications. Other late effects included mild to moderate hearing impairment and dialysis for oliguric renal failure with cisplatin, mild pulmonary function effects with bleomycin, and hemorrhagic cystitis with cyclophosphamide. The results of neoadjuvant chemotherapy followed by primary site RT are more promising compared to chemotherapy alone and are similar to primary RT alone using whole-brain and craniospinal fields, with > 90% survival rates.[16,17] As to the RT volume when using neoadjuvant chemotherapy, most studies have used the primary site with a margin, without causing an excess of relapses.[ 16-18] Two more recent studies that treated the primary site alone after chemotherapy showed an unexpected increase in local relapse.[19,20] RT doses in this setting have ranged from 24 to 50 Gy. The Japanese have shown excellent local control after administering 24 Gy; however, the French had to increase their dose to 40 Gy after a pilot study using 30-Gy primary site RT revealed an increased number of local relapses.[16,17] It seems reasonable to tailor the dose according to response to chemotherapy, as in the Mayo Clinic study, where a complete response (CR) after induction chemotherapy resulted in lowering of the dose from 50 to 30 Gy.[18] The upcoming COG trial should help clarify some of these issues, as the dose of RT will be lowered for patients achieving a CR after induction chemotherapy. Nongerminomatous Germ Cell Tumor
Jubran et al discuss the 20% to 45% overall survival rates associated with the treatment of NGGCT using RT alone. There is no question that the addition of chemotherapy to RT has improved survival rates in these patients.[9,16,18] With regard to RT after neoadjuvant chemotherapy, what is the appropriate volume for localized NGGCT? Controversy exists as to whether the entire craniospinal axis needs to be treated. The German Cooperative Trial MAKEI 89 found a better overall survival in patients receiving craniospinal irradiation compared to those treated with more involved fields.[21] A Japanese multi-institutional study showed no relapses in five treated with craniospinal irradiation, whereas three of eight who did not receive such therapy had a spinal relapse.[22] Others have found involved-field RT to be adequate. Haas-Kogan et al found one spinal relapse in six patients treated with involved-field RT.[9] In the Mayo Clinic experience, seven patients with nonmetastatic NGGCT received involved-field RT, with no relapses.[18] At the Children's Hospital of Denver, five patients with localized alpha-fetoprotein-positive NGGCT were treated with involvedfield RT, without any recurrence.[23] Future investigations are needed to determine the appropriate RT volume for localized NGGCT. With regard to RT dose, there is less variability, as these tumors are not as radiosensitive as pure germinomas; most studies have utilized doses ranging from 50 to 54 Gy.[18,21,22] Conclusions
As with other pediatric tumors, a balance between cure and late toxicity is paramount in determining treatment, especially for intracranial germinoma. For NGGCT, chemotherapy followed by radiotherapy gives the best chance of cure and should be the standard of care. Future studies are under way to answer some of the unanswered questions about germinoma and NGGCT.