CNS non-germinomatous germ cell tumor (NGGCT): Lessons from the recent past

Central nervous system (CNS) germ cell tumors (GCT) are a rare heterogenous group of malignant tumors that commonly arise in children, adolescents, and young adults. These tumors account for approximately 3%-5% of all primary pediatric CNS tumors in North America, with a higher incidence reported in certain parts of Asia.^1,2 CNS GCTs are classified based on their histological appearance and immunohistochemical profile and are divided into germinoma and non-germinomatous germ cell tumors (NGGCTs), with the latter rarer and more challenging to treat. Through numerous prospective clinical trials across the globe, the treatment of CNS GCT has steadily advanced with improvement in outcomes.^3,4 However, a universally accepted standard-of-care treatment for these tumors remains elusive, with varied diagnostic criteria and treatment regimens employed between oncologist in the United States/Canada (Children’s Oncology Group [COG]), Europe (International Society of Pediatric Oncology [SIOP]), and in Asia. Unfortunately, a significant proportion of patients with CNS NGGCT either do not respond to treatment or suffer disease recurrence during/following completion of therapy.⁵ With the noted variance in treatment approaches, coupled with the relative rarity of this disease, there is a paucity of data in the literature describing the pattern of disease recurrence and treatment failure.

To address this gap in the literature, the authors in this study⁶ utilized a transatlantic collaboration to assemble a patient cohort with CNS NGGCT, combining two non-overlapping patient cohorts from five legacy prospective clinical trials and a registry cohort of patients treated with similar approaches. In a disease where large comprehensive datasets are generally lacking, the authors assembled one of the largest cohorts of patients with CNS NGGCT with meaningful clinical follow-up data, which represents one of the major strengths of this research study. Through a thorough interrogation of the patient cohort, several interesting findings were identified. Among others, this manuscript showed that most treatment failures after completion of therapy were metastatic relapses. Notably, when compared with local relapses, patients with metastatic relapses exhibited shorter time-to-recurrence, suggesting a distinct mechanism of treatment resistance/failure between metastatic and local recurrences, as noted by the authors. This study also found that growing teratoma syndrome (GTS) occurs more frequently in younger children, and presents most commonly during therapy and among patients who have normalization of tumor markers.

GTS is an uncommon phenomenon in NGGCT, characterized by the paradoxical growth of the tumor during treatment despite evidence of disease response seen through normalization of tumor markers. First described by Logothetis et al in 1982 among patients with non-CNS GCTs (mainly testicular germ cell tumors), this occurrence is thought to occur in less than 10% of systemic, non-CNS GCT.⁷ While the exact incidence of intracranial GTS is unclear, a recent retrospective study through a large international collaboration reported the frequency to be approximately 5% for CNS GCTs.⁸ It is critically important to note that while GTS has previously been considered as a criterion for PD (progressive disease on response evaluation) in clinical trials (such as COG ACNS0122), it is now clear that true GTS—where there is only mature teratoma on histology without evidence of malignant component on histology—have no real impact on survival outcomes and should not be considered/categorized as treatment failure. The inclusion of this as a local failure in the past may be a reason for reported higher rates of local failure than distant failure.

The authors suggested that disease surveillance of the spine could potentially be negated 1.5 years following completion of therapy—if tumor markers are persistently negative. While this is an intriguing prospect and supported partially by the evidence presented, the wide range of time-to-recurrence in this cohort showed that there is a significant proportion of patients who suffered metastatic recurrence 18 months after completion of treatment (IQR: 8-20.5 months). Importantly, incidences of very late recurrences were noted (range: 6-72 months). Additionally, close to 10% of patients in the metastatic cohort had negative tumor markers at the time of metastatic recurrence. In addition, the requirement for spinal imaging and timing of MRI varied significantly between studies, which may have had an impact on the documentation of patterns of relapse. Taking these data and other historical data into consideration, while ongoing evaluation and optimization of our surveillance protocols after therapy are of critical importance, the evidence of late metastatic recurrence in this patient cohort supports the continued surveillance of the craniospinal axis for patients who have completed therapy for CNS NGGCT.

This study also looked at the controversial topic of the role of spinal irradiation in the treatment of localized CNS NGGCT. While the authors provided some evidence against spinal irradiation, the pattern of relapses seen in the two most recent prospective trials through the COG, however, gives us pause. These two trials (ACNS0122 and ACNS1123) utilized the same chemotherapy backbone for CNS NGGCT, followed by distinct irradiation plans. ACNS0122 administered craniospinal irradiation (36 Gy CSI and 54 Gy to primary tumor bed) to all patients, while ACNS1123 evaluated 30.6 Gy whole ventricular irradiation with 54 Gy to the tumor bed for patients with met specific response criteria. On ACNS0122, the vast majority (76%) of patients had localized disease, with the 3-year event-free survival (EFS) and overall survival (OS) reported at 92% and 94.1%, respectively, for patients with localized disease who achieved a complete response (CR) or partial response (PR). Importantly within this group, most failures were non-metastatic. The ensuing COG study, ACNS1123 Stratum 1, was developed for patients with localized NGGCT who responded well to induction chemotherapy and thought most appropriate for consideration of reduced-intensity radiation therapy (RT).⁹ In this study, 66 of the 107 patients with non-metastatic NGGCT (61.7%) achieved a CR/PR and proceeded to reduced RT, resulting in a 3-year progression-free survival (PFS) and OS of 87.8% and 92.4%, respectively.⁹ While these early outcomes are overall promising, this study was closed prematurely, as it met early stopping rules when nine patients were found to have relapsed. It is important to note that on further retrospective review, two patients with relapse were found to have been ineligible to receive reduced-dose radiation. Nevertheless, the pattern of relapse in this study was particularly concerning, as all relapses involved a spinal component, with six patients having isolated spinal failures, a striking divergence in pattern of failure from ACNS0122. The recently opened COG trial, ACNS2021, aims to determine if the re-addition of spinal canal irradiation will decrease spinal relapses and maintain disease outcomes achieved on ACNS0122.

Lastly, despite the heterogeneity of patient cohort and treatment approach, this manuscript emphasizes the value of international collaboration and of secondary analyses of clinical trials data, especially in diseases where there is a paucity of data. In this case, this manuscript provided critical insights into the pattern of disease relapse in NGGCT and will help inform the design of future clinical trials.