DISCUSSION
In 1998, Satge et al. reported that GCTs make up nearly a third of intracranial tumors in patients with DS, whereas they account for <4% in the general pediatric population.3Though a mechanism for this interaction is unestablished, DS is known to predispose to the development of gonadal GCTs through various mechanisms.3 Few patients with DS and CNS GCTs have since been reported.4 This study includes the largest reported cohort of patients with DS and CNS GCTs and provides insight into the characteristics, management and outcomes of this population.
Genetic and environmental differences influence the occurrence of CNS GCTs, as evidenced by differences in incidence by geographical location. Germ cell tumors account for <5% of all childhood CNS tumors in the United States, Germany and Canada.33-35 In contrast, GCTs account for 2.1-9.4% in Japan and the Far East.36,37 Race and ethnicity also play a role, as an analysis of the SEER database by Poynter et al. showed higher incidence rates of CNS GCTs in patients of Asian/Pacific Islander descent irrespective of gender, tumor location and histology.38 Consistent with these findings, the majority of studies we identified in the literature (22/26) originated from Japan or China. Consistent with prior studies, CNS GCTs were also seen most frequently in male patients.34 Male sex has been strongly associated with pineal involvement, whereas suprasellar involvement has been associated with female sex.39,40In our study, pineal localization showed a slight male predominance (5/9; 55.6%), whereas sellar/suprasellar involvement showed a female predominance (5/8; 62.5%). Overall, CNS GCTs in the general population most commonly occur in the pineal or suprasellar region.1,36,37 Interestingly, the most common site of tumor involvement in our patients was the basal ganglia (13/41; 31.7%). Germ cell tumors arising in the basal ganglia or thalamus are relatively rare, representing only 5-10% of all primary intracranial GCTs.41
Comparable to prior studies, patients in our cohort displayed symptoms for an average of 6.9 months prior to diagnosis.42 In a retrospective review of 30 patients with CNS GCTs, Crawford et al . reported a mean symptom duration of 8.4 months; symptoms including movement disorders, enuresis, anorexia, and psychiatric complaints led to delayed diagnosis (>6 months) in nine patients, whereas headache, nausea, vomiting and visual changes led to earlier diagnosis.42 Similarly, patients in our cohort with symptoms >6 months had movement disorders (n=3) or polyuria/polydipsia (n=3) on presentation.
The Brain Tumor Registry of Japan reported a 5-year OS of 97% for patients with pure germinomas (n=200) from 2001 to 2004; within their database, germinomas made up 64% of all CNS GCTs.37In our study, germinomas accounted for 39% of all tumors and had a poor 3-year OS of 62%. Localized germinomas have historically been managed with WBRT/WVI followed by primary tumor boost, with CSI being reserved for disseminated germinomas.43 More recently, chemotherapy has been incorporated to reduce the late adverse-effects caused by RT.44,45 In an analysis of the SIOP-CNS-GCT-II trial, Calaminus et al. showed that following CR to induction chemotherapy, patients with localized germinomas (n=58) who received 24 Gy WVI without a tumor boost had a 4-year event-free survival of 98%.45 Furthermore, Bartels et al.demonstrated that 18 Gy WVI with 12 Gy tumor boost was associated with an excellent 3-year progression-free survival rate of 94.5.44 In our cohort, most germinoma patients received RT or combined chemotherapy/RT.
In contrast to germinomas, NGGCTs are less common and associated with worse outcomes.46 In an analysis of the European SIOP-CNS-GCT-96 trial by Calaminus et al. , 5-year OS for patients with localized NGGCTs (n =116) was 82%.46 More recent trials have shown better outcomes – In a phase II trial for patients with newly-diagnosed NGGCTs, Goldman et al. demonstrated a 5-year OS of 93% when neoadjuvant chemotherapy was followed by 36 Gy CSI with 54 Gy tumor-bed boost. In comparison, Fangusaro et al.showed that reduced RT (30.6 Gy WVI and 54 Gy tumor-bed boost) resulted in a similar OS of 92.4% at 3 years, though it is important to note that all patients who had treatment failure were in the spine.47 In our cohort, 3-year OS for NGGCTs was 79%.
The mainstay of treatment for teratomas is maximal surgical resection, while adjuvant chemotherapy and RT is recommended for some patients depending on each case and residual disease.43 Huanget al. reported a 5-year OS of 40% in 13 immature teratoma patients treated with maximal possible resection, and no difference in outcomes were found between patients who received chemotherapy and/or RT versus those who did not.48 Three-year OS for teratomas (five immature, one mature, two unspecified) in the present study was comparable at 53%; notably, all three patients who underwent biopsy-only died from tumor progression within one year, whereas all four patients who underwent GTR were alive at follow-up.
The SIOP-96 trial reported grade 3-4 toxicities in 27/83 patients (33%) receiving combined chemotherapy/RT and 26/154 patients (17%) receiving RT only; there were no treatment-related deaths.49 In contrast, while 16/41 patients (39%) in the present study experienced grade 3-4 toxicities, 50% died from their toxicity (n=8). Moreover, five of the 16 germinoma patients (31%) died from treatment-related toxicity, which resulted in a poor 3-year OS of 62%. Trisomy-21 is associated with functional deficiencies in B-cell, T-cell and phagocytic cell systems, resulting in an increased risk of developing and succumbing to various infectious diseases.5,7 In an analysis of the AML-BGM-93 trial for AML, Lehrnbecher et al.found that patients with DS had significantly higher rates of infection-associated fatal complications (5/28; 17.9%) than children without DS (15/276; 5.4%).5 Notably, infection-related mortality significantly decreased in the subsequent AML-BFM 2004 trial (3/61; 4.9%), in which a reduced-intensity chemotherapy regimen was utilized.7
Craniofacial abnormalities, increased secretions and hypotonia associated with DS contribute to higher rates of upper-airway obstruction, particularly after the induction of anesthesia.50 After tonsillectomy or adenoidectomy surgeries, Goldstein et al. found that children with DS more frequently required airway management or observation in the pediatric intensive care unit compared to controls (25% versus0%).50 Finally, patients with DS are at a particular risk of developing Moyamoya due to cranial RT, potentially because several proteins encoded on chromosome 21 including superoxide dismutase 1, interferon gamma receptor and cystathionine beta-synthase affect arterial physiology as well as DNA repair.6
Our study has multiple limitations; despite being the largest cohort of patients with DS and CNS GCTs, the small number of patients limits our conclusions regarding treatment approach. Moreover, our reliance on multiple institutions limits the consistency of pathology or radiology reports, as well as tumor marker levels. The inconsistency of data reporting by previously published studies also limits the thoroughness of our retrospective cohort.