Chemotherapy plus focal radiation therapy for localized intracranial germinoma: How little is enough?

Intracranial germinomas are exquisitely chemo- and radiosensitive tumors, and excellent recurrence-free survival can be achieved with craniospinal irradiation and focal boost to the tumor site, even in patients with metastatic disease.¹ For the majority of patients, who present with (bi-)focal, non-disseminated disease involving the suprasellar and/or pineal region, successive studies have shown that more limited radiotherapy (RT) is sufficient. A cure rate of >90% with acceptable long-term morbidity can be accomplished with 21-24 Gy whole-ventricular irradiation (WVI) with tumor boost to 40-45 Gy.² More recent efforts have aimed to further reduce RT with neoadjuvant chemotherapy, thereby reducing long-term morbidity. Preliminary reports from the SIOP CNS GCT II trial suggest that 24 Gy WVI without boost may be sufficient therapy for localized germinomas with complete imaging response post-chemotherapy.³

In the current issue of Neuro-Oncology, the authors present the results of a Children’s Oncology Group (COG) phase II, response-based RT dose-reduction trial of 18 Gy WVI and boost of 12 Gy to the tumor bed in pediatric patients with localized germinoma who achieved a complete response to a simplified and less-toxic 2-drug chemotherapy regimen consisting of four cycles of carboplatin and etoposide.⁴ This combined chemo + WVI approach aimed to improve quality of life by reducing RT-associated late effects without compromising recurrence-free survival.

The primary objective was to determine the 3-year progression-free survival (PFS) rate compared to a historical cohort. Patients were simultaneously enrolled on the COG ALTE07C1 study (NCT00772200) to longitudinally assess cognitive, social, and behavioral functioning. Multiple domains including processing speed, attention/concentration, and estimated IQ based on Block Design and Vocabulary were assessed using the age-appropriate Wechsler Intelligence Scales.

As anticipated, the simplified chemotherapy regimen was well tolerated and yielded a high response rate, with complete response (CR), partial response (PR), and stable disease (SD) in 62, 30%, and 8% of patients, respectively. After second-look surgery in a subset of patients with PR or SD, the CR rate increased to 68%. Among all eligible patients, only 74 (54%) were evaluable for the primary study objective, due to a high rate of discordant central radiology review and incomplete tumor marker assessment prior to RT. Of these, 5 patients underwent second-look surgery, demonstrating nonviable tumor and/or mature teratoma.

Outcome analysis revealed overall excellent results with the protocol treatment strategy, with a total of only 8 relapses among the 137 eligible patients. Nevertheless, for the primary study aim, the study did not meet the objective of at least 95% 3-year PFS, which was derived from historical controls. However, in contrast to the historical control cohort, the study was designed to count patients who were lost to follow-up or withdrew consent as failures. In retrospect, this appears to have “set the bar” unrealistically high, with 4 patients lost to follow-up and 2 patients who withdrew consent contributing toward the total number of 10 failures on the primary study objective. Using the Kaplan-Meier approach where patients lost to follow-up are treated as censored, the study results indicate excellent outcomes for the primary aim with protocol therapy, with 3-year estimated PFS and overall survival (OS) of 94.5 ± 2.7% and 100%, respectively. Results for the 16 evaluable patients with PR who received 24 Gy WVI were excellent as well, with 3-year PFS and OS of 93.7 ± 6.1% and 93.7 ± 6.1%, respectively.

Notably, 3 of the 8 relapses (37.5%) occurred along a surgical tract (ie, biopsy/endoscopy tract, external ventricular drain), raising the question whether surgical tracts should be included in the radiation field, along with prior reports.^5,6 Careful further analysis of surgical tracts and RT fields from this and other cooperative group trials may help shed further light on this important issue.

While patients with residual disease more than 1.5 cm after chemotherapy were required to undergo surgical resection, none of the 11 study patients who underwent second-look surgery showed any viable germ cell tumor elements. Furthermore, no recurrences were reported among 18 patients who were taken off study prior to RT due to inability/unwillingness to undergo second-look surgery and/or physician discretion. These findings are in line with retrospective data from Asia⁷ and the SIOP working group,³ indicating that residual disease post-therapy was not associated with a worse prognosis. Therefore, the neurosurgical risks associated with resection of residual lesions (aside from growing teratomas) must be carefully considered.

Despite the relatively small sample sizes, neurocognitive findings indicated improved performance and IQ for those patients treated with 18 Gy vs 24 Gy WVI, and long-term follow-up data will be required to confirm the promising preliminary cognitive benefits of irradiation reduction strategies.

The authors should be commended given the formidable challenges associated with designing and completing a therapy-reduction trial in a rare and clinically heterogeneous disease entity. As they point out, future studies should incorporate real-time staging review, including imaging and tumor markers, at baseline and prior to RT, because even a small number of patients with incomplete (re-)staging have the potential to skew results. This issue was highlighted in the recently published COG therapy-reduction trial for patients with localized non-germinomatous germ cell tumors.⁸

Despite the failure to meet the primary non-inferiority endpoint given the perhaps overly conservative statistical design, the encouraging results from this study make significant strides in identifying less toxic therapies for patients with localized intracranial germinoma. While further studies will be needed to define optimal treatment for this population, the overall excellent outcome for patients treated with the protocol strategy suggests that the simplified, less toxic chemotherapy regimen followed by response-based reduced-dose WVI could be considered a reasonable treatment approach outside of a clinical trial, with the potential of meaningfully reduced long-term morbidities. While the study was designed for the pediatric population, the relevance of the findings also translates to young adult patients with localized intracranial germinoma. Pending further in-depth analysis of the patients who experienced a relapse, treating radiation oncologists may consider the inclusion of surgical tracts in the radiation field.

Acknowledgments

This editorial is the sole product of the authors, and we thank Dr. Suzanne Wolden for critical review. No third party provided input or supported its writing.

Funding

None.

Conflict of interest statement. S.F.S. declares consulting agreements with AstraZeneca (personal fees received) and QED Therapeutics. M.A.K. declares consultant agreements with AstraZeneca, Bayer, CereXis, Recursion Pharma, QED Therapeutics, and Cardinal Health (personal fees received).

References